Biological control of botrytis bunch rot in organic wine grapes with the yeast antagonist Candida sake CPA-1




The aim of this research was to confirm the efficacy of the yeast antagonist Candida sake CPA-1 in suppressing botrytis bunch rot development, in an organic vineyard under Mediterranean conditions for two seasons, and compare its performance with that of two biologically based products currently registered for botrytis bunch rot control in New Zealand. In 2009, treatments applied were: commercial formulations of Ulocladium oudemansii (BOTRY-Zen®) and chitosan (ARMOUR-Zen®), C. sake CPA-1 combined with the fatty acid-based additive Fungicover® and combinations of these products. All treatments were applied six times between early flowering and harvest and compared with an unsprayed control. In 2010, the treatments focused on C. sake and Fungicover and the number of applications was reduced from six to four. The population dynamics of U. oudemansii and C. sake were measured and wine quality tests were carried out in both seasons. Disease control achieved by C. sake treatments in 2009 were comparable to those achieved by BOTRY-Zen and ARMOUR-Zen. Applications of C. sake plus Fungicover between flowering and harvest significantly (< 0·05) reduced botrytis bunch rot incidence and severity by 64% and 90%, respectively, compared with the untreated control in 2009, and by 67% and 89%, respectively, in 2010. Treatments did not adversely affect wine quality parameters after treated grapes were processed. Candida sake consistently provided effective control of botrytis bunch rot in grapes under different meteorological and disease pressure conditions, thereby improving its potential for future commercial applications.


The necrotrophic pathogen Botrytis cinerea is a filamentous fungus of the Sclerotiniaceae family that is able to infect a broad range of hosts (Holz et al., 2004). In grapevines it is responsible for botrytis bunch rot or grey mould and in many temperate regions it is the most important fungal disease that affects grape production before harvest and wine quality postharvest (Elmer & Reglinski, 2006). Conventional approaches to botrytis bunch rot control have focused on synthetic fungicide applications. However, sole reliance upon this approach is not sustainable because of the emergence of fungicide resistance in vineyard populations of B. cinerea (Leroch et al., 2011) and the adverse effects of synthetic pesticides on environmental and human health (Komárek et al., 2010). Increasing consumer demand for no detectable pesticide residues in wine has forced many producers to restrict the application of synthetic fungicides to the early part of the growing season, thereby reducing the risk of residue being detected in the wine (Elmer & Michailides, 2004). Unfortunately, this strategy is flawed because rapidly ripening berries in the post-véraison period are highly susceptible to infection by B. cinerea (Hill et al., 1981) and in the absence of a suitable protectant, significant crop losses may occur. Substituting synthetic fungicides in the mid- and late season with commercially available biologically based products has been reported to provide acceptable control of botrytis bunch rot in grapes, indicating that there is potential to integrate early season synthetic fungicides with biologically based products after bunch closure (Parry et al., 2011).

In line with low pesticide strategies, organic viticulture continues to increase globally. The area of grapes grown organically in Catalonia increased by 140% between 2008 and 2010, while in Spain an overall increase of 5% of the area devoted to organically grown grapes was registered in the 2009–2010 period (MARM, 2010), further highlighting the importance of this sector of production in the future.

In organic viticulture, botrytis bunch rot control is usually carried out through cultural methods and canopy management, while other organically acceptable methods include spray applications of bentonite clays, copper-based formulations, compost teas or plant extracts (Jacometti et al., 2010). Botrytis bunch rot control based upon these strategies may be variable or have undesirable consequences on the aromatic compounds of some varieties (Jackson, 2008). These issues further highlight the urgent need for new organically acceptable full-season treatments for botrytis bunch rot control to complement canopy management and cultural practices.

A diverse range of alternative strategies has been evaluated by many research groups worldwide with special emphasis on natural products, elicitors of host defences and antimicrobial antagonists. Recent reviews of these alternative strategies include those by Elmer & Reglinski (2006) and Jacometti et al. (2010). A natural product that has shown promise against B. cinerea in grapes is chitosan: a naturally occurring polysaccharide derived from chitin. Efficacy studies on grapes have focused on postharvest treatment of table grapes (Xu et al., 2007; Romanazzi et al., 2009). Relatively fewer studies have focused on preharvest field applications to wine grapes, with the exception of Mehofer et al. (2009) and Reglinski et al. (2010). A chitosan-based product has been commercialized in New Zealand (ARMOUR-Zen®) with a label claim for botrytis bunch rot control in wine grapes (Reglinski et al., 2010). Biological control of B. cinerea in grapes using naturally occurring microbial antagonists has been widely studied and several products have been developed and commercialized (Elmer & Reglinski, 2006; Jacometti et al., 2010). An isolate of Ulocladium oudemansii was formulated and commercialized (BOTRY-Zen®), showing efficacy against botrytis bunch rot when applied during the early part of the growing season in New Zealand vineyards (Reglinski et al., 2010; Wurms et al., 2011).

Neither ARMOUR-Zen nor BOTRY-Zen has been evaluated under the hot and dry environmental conditions that occur during parts of the growing season in Catalan vineyards.

Yeasts have also been evaluated as potential biological control agents (BCAs) for B. cinerea (Elad & Stewart, 2004). The most important genera include Pichia, Candida, Metschnikowia (Jacometti et al., 2010) and, more recently, Saccharomyces cerevisiae (Nally et al., 2012). In field and laboratory studies, pre- and postharvest applications of Candida spp. have given effective control of B. cinerea in a wide range of crops; for example, C. oleophila (El-Neshawy & El-Morsy, 2003) and C. guillermondii (Scherm et al., 2003) when applied to apples and C. saitoana (El Ghaouth et al., 2003) for postharvest disease control in apples and citrus.

Candida sake CPA-1 is a yeast strain formulated and optimized by the postharvest pathology research group at IRTA Lleida. It was effective against postharvest B. cinerea in pome fruits (Torres et al., 2006), suggesting its potential application against B. cinerea on other crops, such as grapes.

A previous study (Cañamás et al., 2011) evaluated the effectiveness of C. sake CPA-1 against botrytis bunch rot applied under dry Mediterranean conditions (Catalonia, Spain) in conventional vineyards. Heat-adapted and non-adapted fresh cells were tested in different formulations with an isotonic solution or with the additive Fungicover. Fungicover was already known to be a beneficial additive to C. sake, improving the persistence of the BCA on the host and its efficacy to levels comparable to that of fungicide treatment. Thus, the application of C. sake CPA-1 and Fungicover as a treatment mixture was regarded as an interesting strategy to be further investigated in field trials.

This study was undertaken over two consecutive growing seasons in the Lleida winegrowing area (Catalonia, Spain). The aim was to confirm the effectiveness of field applications of treatments with C. sake CPA-1 plus Fungicover against botrytis bunch rot in organic vineyards, at different doses and numbers of applications, and compare their efficacy with that of other registered biological products. Treated vines were also processed into wine using microvinification techniques, to evaluate possible adverse effects of BCA treatments on wine quality.

Materials and methods

Microbial antagonists and natural products

Candida sake strain CPA-1 was isolated from the surface of apples at the University of Lleida-IRTA research centre and was deposited in the Colección Española de Cultivos Tipo (CECT-10817) at the University of Valencia, Burjassot, Spain. Cell production and formulation prior to field application was carried out according to methods described by Cañamás et al. (2011).

Ulocladium oudemansii isolate HRU3 was applied during flowering and up to pre-bunch closure using the formulated product BOTRY-Zen® (Botry-Zen Ltd). Chitosan was applied using the formulated product ARMOUR-Zen® (Botry-Zen Ltd). All treatments with BOTRY-Zen and ARMOUR-Zen included the wetting agent Mojante-Inagra (Sipcam Inagra S.A.) at 0·5 mL L−1. ARMOUR-Zen treatments were also supplemented with 1·05 g L−1 sodium bicarbonate in order to adjust the final solution to pH 7, because studies indicated that this adjustment improved B. cinerea suppression in vitro (T. Reglinski, Plant and Food Research, New Zealand, personal communication).

Fungicover (BioDúrcal S.L.) is a commercial formulation of derivatives of fatty acids in an aqueous-alcoholic solution and previous research showed that this additive improved C. sake survival on grape host tissues (Cañamás et al., 2011). Its use is authorized in Spain for conventional and organic agriculture by AGROCOLOR S.A., authorized agency under the Council Regulation (EEC) No. 2092/91, as a protector for preventing fungal diseases on a variety of crops (BioDúrcal, 2010).

Experimental field site (2009 and 2010)

Field experiments in both growing seasons were carried out in a commercial organic vineyard certified by the Catalan committee for organic agriculture production (CCPAE) and located in the Designation of Origin Costers del Segre (Lleida, Spain). The experimental site was located in a part of the vineyard with a history of botrytis bunch rot and trained into a bush system. The grape cultivar used was Macabeu, characterized by large and very compact clusters and thick-skinned berries (Fuster, 2006). These characteristics are reported to increase susceptibility to B. cinerea (Fermaud et al., 2001). Vines were sprayed with organically acceptable treatments throughout both growing seasons by the vineyard manager for pests and diseases. In 2009 these treatments consisted of three applications of 99% sulphur (w/w) prior to flowering and one more application of 99% sulphur (w/w) plus 98% silicon (w/w) at pre-bunch closure; in addition, one application of Bacillus thuringiensis var. kurstaki was made at véraison. In 2010, treatments included one application of 99% sulphur (w/w) prior to flowering, two applications of 60% sulphur plus 4% copper oxychloride at pre-bunch closure, one application of 38% copper oxychloride (w/v) at pre-bunch closure and two applications of B. thuringiensis var. kurstaki during véraison. All products were applied at the doses and application rates recommended by their manufacturers.

Experimental design

Plots were distributed in a completely randomized block design, with four replicates per treatment. Each plot consisted of seven vines; the first and last vines were used as buffer vines, the next two to monitor BCA population dynamics and the last three to measure botrytis bunch rot development.

All treatments were applied at key vine phenological stages (1–5% flowering, 80% flowering, pre-bunch closure, véraison, 21 and 7 days before harvest) using a motorized backpack sprayer (model WJR2225; Honda Motor Company Ltd) with a 1-mm nozzle and 15 bar pressure, spraying each grape bunch until run-off.

Bunch rot management programmes

The botrytis bunch rot management programmes used in 2009 are detailed in Table 1.

Table 1. Biologically based botrytis bunch rot control programmes applied to Macabeu wine grapes and botrytis bunch rot at harvest in an organic vineyard in 2009
TreatmentaVine phenologyBotrytis bunch rot at harvestb
1–5% Flowering80% FloweringPre-bunch closureVéraison21 days before harvest7 days before harvestIncidence (%)Severity (%)
  1. a

    Control: untreated; CS + FC: Candida sake at 5 × 107 colony-forming units (CFU) mL−1 + Fungicover at 50 g L−1; CS low + FC: C. sake at 1 × 107 CFU mL−1 + Fungicover at 50 g L−1; AZ: chitosan at 1·44 g L−1 + wetting agent at 0·5 mL L−1 + NaHCO3 at 1·05 g L−1; BZ: Ulocladium oudemansii at 2·5 × 106 CFU mL−1 + wetting agent at 0·5 mL L−1; FC: Fungicover at 50 g L−1.

  2. b

    All values are means of four replicates. Mean values with the same letter are not significantly different (= 0·05) according to LSD Student's t-test.

CS + FCCS + FCCS + FCCS + FCCS + FCCS + FCCS + FC28·5c0·8b
CS low + FCCS low + FCCS low + FCCS low + FCCS low + FCCS low + FCCS low + FC43·5b1·9b
BZ-(CS + FC)BZBZBZCS + FCCS + FCCS + FC39·0bc1·4b
(CS + FC)-AZCS + FCCS + FCCS + FCAZAZAZ30·0bc1·2b

Final concentrations of the active ingredients applied were: chitosan 1·44 g L−1, U. oudemansii 2·5 × 106 colony-forming units (CFU) mL−1 and C. sake 5 × 107 CFU mL−1. An additional treatment tested a reduced concentration of C. sake, 1 × 107 CFU mL−1, in order to improve the potential economics of C. sake as a new BCA for application in vineyards. Fungicover® was applied at a dose of 50 g L−1 for all C. sake treatments.

In 2010, nine botrytis bunch rot control programmes were applied (Table 2), focusing on C. sake CPA-1 at 5 × 107 CFU mL−1 combined with Fungicover at different doses (50 or 25 g L−1) and different application timings. Treatment applications were reduced to four or two, in order to evaluate whether more cost-effective spray programmes were achievable. Application methods and formulations of C. sake CPA-1 and Fungicover were the same as in 2009.

Table 2. Candida sake CPA-1 and Fungicover botrytis bunch rot control programmes applied to Macabeu wine grapes and botrytis bunch rot at harvest in an organic vineyard in 2010
TreatmentaVine phenologyBotrytis bunch rot at harvestb
80% FloweringPre-bunch closureVéraison21 days before harvestIncidence (%)Severity (%)
  1. a

    Control: untreated; CS: C. sake at 5 × 107 colony-forming units mL−1; FC: Fungicover at 50 g L−1; FC25: Fungicover at 25 g L−1; ×4: four applications from flowering to harvest; ×2: two applications in early season (80% flowering and pre-bunch closure).

  2. b

    All values are means of four replicates. Mean values with the same letter are not significantly different (= 0·05) according to LSD Student's t-test.

CS + FC × 4CS + FCCS + FCCS + FCCS + FC29·5e2·3b
FC × 4FCFCFCFC55·0bcd5·6b
CS + FC × 2CS + FCCS + FC44·0d3·2b
FC × 2FCFC59·5bc4·6b
CS + FC25 × 4CS + FC25CS + FC25CS + FC25CS + FC2546·5d3·3b
FC25 × 4FC25FC25FC25FC2559·5bc3·8b
CS + FC25 × 2CS + FC25CS + FC2548·5cd3·9b
FC25 × 2FC25FC2562·5bc4·1b

Botrytis bunch rot assessment

At harvest (14 September 2009 and 15 September 2010) botrytis bunch rot incidence and severity were assessed on 50 bunches per replicate plot. Twenty-five bunches were selected from each side of the row, avoiding vines used for population dynamics studies and buffer vines. Incidence was measured as the percentage of bunches with visual B. cinerea infection symptoms. Bunch rot severity was measured as the percentage of B. cinerea-infected berries per bunch.

Meteorological data

Temperature (T) and relative humidity (RH) were logged at hourly intervals in both seasons using a weather station (Decagon Services Inc.) placed at the experimental field site.

Population dynamics of biological control agents (2009 and 2010)

Populations of C. sake or U. oudemansii were measured in the 2009 and 2010 seasons by sampling treated flowers or berries (i) once the tissue surface was dry, just after treatment application, and (ii) before the next treatment application. Sampling of host tissues was carried out using sterile clippers with each plot sample placed directly into 50-mL sterile plastic tubes (Corning Inc.) that were then cool-stored prior to tissue processing on the same day. At flowering, BCA populations were recovered from 2 g floral organs from eight bunches per plot, collecting two groups of approximately 10 flowers per bunch. Samples where then immersed in 20 mL phosphate buffer in 250-mL conical flasks. At pre-bunch closure 40 pea-sized berries from 20 bunches per plot were weighed and then immersed in 50 mL phosphate buffer in 250-mL conical flasks. After véraison, 20 berries from 10 bunches were also weighed and then immersed in 50 mL phosphate buffer. All sampled tissues were shaken for 20 min at 150 rpm on a rotary shaker and then sonicated for 10 min in an ultrasonic bath (JP Selecta S.L.). After serial dilutions, 100-μL aliquots were plated onto NYDA plates supplemented with streptomycin sulphate (0·5 g L−1) for C. sake population estimates, or on Bengal rose agar (Biokar Diagnostics) plates amended with chloramphenicol (0·1 g L−1) for estimates of U. oudemansii populations. Duplicate plates were incubated in the dark at 25°C, and colony counts measured after 48 h (for C. sake) and after 6 days (for U. oudemansii). For both BCAs, colonies were visually recognized based on their morphological characteristics. Data were collected as CFU mL−1 and expressed as CFU g−1.

Microvinification of C. sake-treated grapes (2009 and 2010)

Grape bunches presenting similar ripeness and without bunch rot symptoms were visually selected at harvest and then transported to the winery for processing. In 2009, 10 kg grapes were collected from each treatment replicate plot, as fermentation was performed in small batches of 1 L. In 2010, 30 kg were also sampled from each of the four replicates of the evaluated treatments, conducting microvinifications in 20-L tanks. In both years, the grapes harvested from four replicate plots of the field experiment were redistributed into three replicate batches for the fermentation evaluation.

Grapes were crushed and pressed, sulphite added (40 mg L−1) and the crushed juice allowed to settle overnight. After racking off the lees, the clear juice was inoculated with selected wine yeast (Lallemand EC1118). Fermentations were monitored by measuring density and quantity of residual sugar. Once the sugars were fermented, the wine was racked off the lees, cold-stabilized and then filtered.

Key wine quality parameters (ethanol, volatile acidity, reducing sugars, pH, titratable acidity and malic acid) were analysed following OIV methods (O.I.V., 2008). A descriptive sensory analysis was also performed by a trained panel (12 people) with samples treated with C. sake and Fungicover in 2010, following the standards of the International Organization for Standardization (ISO, 2008); wine samples were tasted in triplicate in tastes held on different days. Each tasting was conducted in individual tasting booths at room temperature (22°C). In each case, wines (50 mL) were served in coded, tulip-shaped wine glasses covered with plastic lids. Samples were presented in random order. Still mineral water was available for rinsing between wines.

Statistical analysis

Analysis of variance was performed using jmp®8 statistical discovery software (SAS Institute Inc.), for all data sets. Significant treatment differences were determined using LSD test (= 0·05). Orthogonal contrasts were used in the analysis of botrytis bunch rot incidence in the 2010 season to gain greater understanding of the overall treatment effects. CFU data for BCA population counts were log-transformed prior to anova to stabilize the variance.


2009 field studies

Botrytis bunch rot incidence at harvest was high (80·0%) in untreated plots, and average bunch rot severity was 8·2% (Table 1). All treatments significantly reduced (< 0·05) botrytis bunch rot incidence by 45% (CS low + FC) to 64% (CS + FC), and severity from 73% (AZ) to 90% (CS + FC) compared with the untreated control. When the low rate of C. sake (CS low + FC, 1 × 107 CFU mL−1) was applied, botrytis bunch rot incidence increased and was significantly higher than with the standard (high) rate of C. sake (5 × 107 CFU mL−1, CS + FC). There were no significant differences (< 0·05) in disease severity among applied treatments.

2010 field studies

Incidence and severity of botrytis bunch rot in the untreated control was higher in 2010 (89·5% and 21·7% respectively; Table 2) than in 2009. All treatments significantly reduced (< 0·05) botrytis bunch rot incidence by 30% (FC25 × 2) to 67% (CS + FC × 4) and severity by 74% (FC × 4) to 89% (CS + FC × 4) compared to the untreated control. The most effective treatment was CS + FC × 4, which reduced botrytis bunch rot incidence from 89% (untreated control) to 30%. Botrytis bunch rot severity in treated plots was always lower than 5·6%, with no significant differences among treatments.

Overall, treatments with C. sake had significantly (< 0·001) less botrytis bunch rot incidence than treatments with Fungicover only (Table 3). Treatments including the high dose of Fungicover (50 g L−1) presented significantly (< 0·05) lower incidence than those applied at the lower dose (25 g L−1). Contrasts analysis also indicated that overall botrytis bunch rot control was significantly better (< 0·05) when four applications of C. sake and Fungicover were applied during the growing season, compared with two applications. None of the interactions among C. sake presence in treatments, Fungicover dose and number of applications were significant (< 0·05), as expressed in Table 3.

Table 3. Orthogonal contrasts testing treatment effects on botrytis bunch rot incidence in 2010 field experiment. Coefficients of equal value indicate pooled levels of the factor, coefficients of opposite sign indicate factor levels to be contrasted and zero indicates excluded factor levels
Effect testbTreatmentsa P
ControlCS + FC × 4CS + FC × 2CS + FC25 × 4CS + FC25 × 2FC × 4FC × 2FC25 × 4FC25 × 2
  1. a

    Control: untreated; CS: Candida sake at 5 × 107 colony-forming units mL−1; FC: Fungicover at 50 g L−1; FC25: Fungicover at 25 g L−1; ×4: four applications from flowering to harvest; ×2: two applications in early season (80% flowering and pre-bunch closure).

  2. b

    C. sake: comparing treatments with or without C. sake CPA-1; FC dose: comparing treatments with Fungicover at 25 or 50 g L−1; number of applications: comparing treatments with two or four applications; interactions among these effects are marked by *.

C. sake 01111−1−1−1−1<0·0001
FC dose011−1−111−1−10·016
C. sake*FC dose011−1−1−1−1110·242
Number of applications01−11−11−11−10·046
C. sake*Number of applications01−11−1−11−110·451
FC dose*Number of applications01−1−111−1−110·242
C. sake*FC dose*Number of applications01−1−11−111−10·353

Meteorological data

Temperature and relative humidity data between each growth stage for both seasons are summarized in Table 4. The 2009 season was characterized by relatively constant conditions before véraison, with warm average temperatures and RH over 59% from late flowering to pre-bunch closure, followed by a very hot and dry period from véraison to before harvest. In 2010, cooler temperatures and higher RH were measured than in the previous year, except for the period from late flowering to pre-bunch closure, in which the average RH registered was the lowest of either season. From véraison to harvest, the 2010 season was consistently cooler and more humid than the 2009 season.

Table 4. Vine growth stages and meteorological data from a commercial organic vineyard in Catalonia in 2009 and 2010 seasons
Vine phenology20092010
DatesAverage RH (%)aAverage temperature (°C)DatesAverage RH (%)Average temperature (°C)
  1. a

    Average temperature and RH values are means of the daily mean values of each variable during the period between dates.

  2. b

    Harvest date in the 2010 season was brought forward by the vineyard manager because of the high levels of botrytis bunch rot registered.

5% flowering8 June  12 June  
  54·522·4 64·716·3
80% flowering15 June  21 June  
  59·322·5 49·724·2
Pre-bunch closure2 July  12 July  
  63·822·9 67·722·7
Véraison11 August  18 August  
  54·725·1 67·222·9
21 days before harvest25 August  1 September  
  68·921·3 71·320·1
Harvest14 September  15 Septemberb  

Population dynamics of biological control agents

The population dynamics of C. sake and U. oudemansii over time are summarized in Figures 1 and 2. CFU counts for each BCA treatment were high immediately after application, then declined over time. Populations of C. sake in the 2009 season (Fig. 1) were similar when comparing the different C. sake treatments, and ranged from 106 to 104 CFU g−1 sample when the concentration of the yeast in the spray suspension was 5 × 107 CFU mL−1. When the lower concentration of C. sake was applied (CS low + FC), recovered populations were approximately one log unit lower at all sample times, compared with the application of the C. sake at the high rate. However, these differences in populations were significant (< 0·05) only at five sample times: 5% flowering, prior to 80% flowering, pre-bunch closure, 7 days before harvest and harvest (data not shown).

Figure 1.

Population dynamics of two biological control agents Candida sake CPA-1 and Ulocladium oudemansii, applied to Macabeu wine grapes in a commercial organic vineyard in 2009. Early season applications were carried out between 5% flowering and pre-bunch closure, late-season applications were applied from véraison to harvest. Solid lines represent C. sake and dotted lines represent U. oudemansii colony-forming units (CFU). Treatments were: CS + FC (■): C. sake (CS) at 5 × 107 CFU mL−1; CS low + FC (×): C. sake at 1 × 107 CFU mL−1; BZ-AZ (●): three applications of U. oudemansii (BZ) at 2·5 × 106 CFU mL−1 early season and then three applications of chitosan (AZ) at 1·44 g L−1 (véraison to harvest); BZ-(CS + FC): three applications of U. oudemansii at 2·5 × 106 CFU mL−1 (Δ) (early season) and then three applications of C. sake at 5 × 107 CFU mL−1 (▲) (late season); (CS + FC)-AZ (○): three applications of C. sake at 5 × 107 CFU mL−1 (early season) and then three applications of chitosan at 1·44 g L−1 (late season). For each application, chitosan and U. oudemansii were applied with a generic wetting agent at 0·5 mL L−1, chitosan treatments included 1·05 g NaHCO3 L−1 and C. sake was applied with the additive Fungicover (FC) at 50 g L−1. Flower or berry samples were taken after spraying and again just prior to the next spray application. CFU values are per gram of tissue sampled and were log-transformed. Values are the means of four replicates.

Figure 2.

Population dynamics of Candida sake CPA-1 applied in 2010 field experiments to control botrytis bunch rot in Macabeu wine grapes. Treatments were C. sake (CS) at 5 × 107 colony-forming units (CFU) mL−1 with Fungicover at 50 g L−1 [CS + FC × 4 (■) and CS + FC × 2 (▲)], or Fungicover at 25 g L−1 [CS + FC25 × 4 (□) and CS + FC25 × 2 (Δ)]. Treatments labelled × 4 were applied four times from flowering to harvest; those labelled × 2 were applied at 80% flowering and pre-bunch closure. Flower or berry samples were taken after spraying and just before the next spray application. CFU values obtained per gram of sample tissue were log-transformed and values shown are means of four replicates.

The C. sake populations followed the same pattern after véraison, except the (CS + FC)-AZ treatment that did not receive any additional C. sake treatment. Populations on berries decreased gradually in these samples, and at harvest had declined to the point where they could not be detected.

The population dynamics of U. oudemansii, BZ-(CS + FC) and BZ-AZ treatments are shown in Figure 1. The populations of U. oudemansii on grape tissues over time were similar in the BZ-AZ and BZ-(CS + FC) treatment samples. On green flowers, U. oudemansii populations ranged from 4·2 to 4·8 log (CFU g−1) after spray application. Between 5% and 80% flowering, population counts declined by one log unit, whereas between the end of flowering and pre-bunch closure CFU g−1 declined (< 0·05) from 6 to 2 log.

No more BZ applications were applied after that time and U. oudemansii populations declined steadily to the point where they were not detected at harvest. The last application was at pre-bunch closure and populations recovered at this stage were two log units lower than those after flowering treatments, ranging from 1·8 to 2·7 log (CFU g−1) before véraison. After véraison populations decreased gradually in BZ-AZ-treated plots. In contrast, there was a substantial decline in the BZ-(CS + FC) treatment.

In 2010, the population dynamics of C. sake treatments were measured (Fig. 2). The populations of C. sake followed a similar pattern, which was predictable because the BCA concentration was the same for each treatment. No significant differences in C. sake populations were detected (< 0·05) when Fungicover concentration was reduced from 50 to 25 g L−1 in any of the sampling dates (data not shown). After véraison, the C. sake population decreased gradually when only two applications of C. sake were made over the growing season, and were consistently over 102 CFU g−1.

Wine quality parameters

Results of final wine analysis are listed in Table 5. All wines fermented correctly, finishing the sugar, reaching the same alcohol amount and with normal volatile acidity considering the small size of the 2009 fermenters. In 2010 with 20 L fermentation, wines were also finished correctly with low residual sugar, similar alcohol content and low volatile acidity. In 2010, total acidity, pH and malic acid were also measured. Significant differences (< 0·05) were detected in the wine quality parameters evaluated (Table 5). Such differences may be produced by variability in grape ripeness. The results indicated that the vineyard treatments did not affect the normal fermentation of the juice and that the resulting wines showed normal analytical composition. No significant differences were detected among treatments in the fermentation dynamics at any sampling time (data not shown).

Table 5. Wine quality parameters after fermentation of Macabeu grapes using microvinification in 2009 and 2010
TreatmentEthanol (% v/v)Volatile acidity (g L−1)Reducing sugars (g L−1)pHTitratable acidity (g L−1)Malic acid (g L−1)
  1. a

    Control: untreated; CS + FC: (Candida sake at 5 × 107 colony-forming units (CFU) mL−1 + Fungicover at 50 g L−1); CS low + FC (C. sake at 1 × 107 CFU mL−1 + Fungicover at 50 g L−1); AZ: (chitosan at 1·44 g L−1 + wetting agent at 0·5 mL L−1 + NaHCO3 at 1·05 g L−1); BZ: (Ulocladium oudemansii at 2·5 × 106 CFU mL−1 + wetting agent at 0·5 mL L−1); FC: (Fungicover at 50 g L−1).

  2. b

    Control: untreated CS: C. sake at 5 × 107 CFU mL−1; FC: Fungicover at 50 g L−1; FC25: Fungicover at 25 g L−1; ×4: four applications from flowering to harvest; ×2: two applications in early season (80% flowering and pre-bunch closure).

  3. c

    Mean values with the same letter are not significantly different (= 0·05) according to LSD test.

  4. d

    Not assessed.

 CS + FC10·58a0·40a0·53aNANANA
 CS low + FC10·79a0·37a0·47aNANANA
 BZ-(CS + FC)10·20a0·37a0·20aNANANA
 (CS + FC)-AZ10·68a0·30a0·30aNANANA
 CS + FC × 49·39b0·14c0·26b2·77d7·90ab2·50d
 FC × 49·32b0·21a0·13b2·88a7·73b2·73b
 CS + FC25 × 49·35b0·12d0·86a2·85b7·03c1·90d
 FC25 × 48·69c0·19b0·06b2·83b8·23a3·10a
 FC + CS × 29·51ab0·16c0·23b2·84b7·86ab2·63c

In 2010 the grapes were harvested before full maturity and presented higher acidity and lower alcohol levels than usual in wines elaborated in the area. Nevertheless, the wine tasting conducted by a trained panel showed no differences among the wines in colour, aroma and taste (data not shown).


The urgent need for biologically based alternatives for control of B. cinerea in grapes has stimulated much research in the last two decades on a variety of natural products and BCAs. Because C. sake CPA-1 demonstrated effective control of botrytis bunch rot of grapes in conventional vineyards in 2005 and 2006 (Cañamás et al., 2011), this study tested the efficacy of C. sake treatments in organic vineyards, evaluating different application timings as well as BCA and additive doses. Comparison with other registered products and integration of these strategies was also evaluated.

High botrytis bunch rot reductions achieved by full-season C. sake and Fungicover treatments in both seasons coincide with those reported by Cañamás et al. (2011) in 2006 (90% and 95% reductions in incidence and severity, respectively). The application of a low-rate C. sake treatment (CS low + FC) also significantly reduced both the incidence and severity of botrytis bunch rot, indicating that reduced rates of this BCA are a possible strategy in the future. However, the incidence of botrytis bunch rot was significantly higher in this treatment than with the full rate of C. sake, indicating that there may be some loss of field efficacy when a lower dose is applied, probably related to reported lower population counts in CS low + FC plots. Therefore, the standard rate of 5 × 107 CFU mL−1 is the preferred BCA concentration for effective and consistent control of botrytis bunch rot in grapes.

In the 2010 studies, once C. sake efficacy was confirmed to be similar to that of other commercialized products, the research focus was field evaluation of fewer C. sake applications per season and reduced rates of Fungicover. Botrytis bunch rot disease pressure in the vineyard was very high in 2010 compared with 2009, favoured by more suitable meteorological conditions for B. cinerea infection. Despite greater disease pressure, four applications at key phenological growth stages were sufficient for effective botrytis bunch rot control. Two applications of C. sake plus Fungicover before véraison were less effective than full-season treatments. Nonetheless, registered efficacy by treatments applied twice was also elevated, providing low-input alternatives for disease control, especially if climatic conditions after véraison are not highly conducive for B. cinerea development. Consistent botrytis bunch rot control under a wide range of conditions in 2006 (Cañamás et al., 2011), 2009 and 2010 represents a significant new advance, because previous researchers have questioned whether BCAs can significantly reduce botrytis bunch rot infection when conditions are highly conducive to disease development (Metz et al., 2002). Moreover, observed efficacy was comparable to other registered products in 2009, corroborating the potential of this BCA for commercial application.

When Fungicover was applied as a standalone product, botrytis bunch rot incidence and severity were significantly reduced. The precise mechanism of action of Fungicover against B. cinerea is poorly understood, but the manufacturers claim that infection is reduced by a physical barrier that this product forms on fruit surfaces after application (BioDúrcal, 2010). The effect of the BCA compared with treatments with Fungicover alone was not significant in 2009. However, the evaluation of C. sake and Fungicover treatments carried out in 2010 showed a highly significant BCA effect (< 0·001). The effect of Fungicover on C. sake survival has been previously reported (Cañamás et al., 2011). This additive also protects the yeast antagonist from exposure to external factors, while a certain nutritive effect has also been observed (C. Calvo-Garrido, N. Teixidó, unpublished data), which may favour the mode of action of C. sake based on nutrient competition. However, the interaction between C. sake and Fungicover dose was not significant and BCA survival was similar for both doses of the additive. Therefore, despite the high efficacy of treatments with C. sake plus a high dose of Fungicover, a synergistic effect is unlikely to occur.

Six applications of chitosan during the growing season resulted in botrytis bunch rot control equivalent to that with C. sake, and comparable to that reported in New Zealand vineyards over two growing seasons (Reglinski et al., 2010). The treatment with U. oudemansii and then chitosan (BZ-AZ) also reduced botrytis bunch rot incidence and severity, achieving similar reduction levels to those reported by Reglinski et al. (2010) and other studies applying U. oudemansii alone or combined with synthetic fungicide programmes in New Zealand (Elmer et al., 2005). The results reported here and in New Zealand are in contrast with those of Mehofer et al. (2009), who field-tested both BOTRY-Zen and ARMOUR-Zen for three seasons on Riesling grapes in Germany. These findings indicate a significant interaction between the performance of biologically based products, climatic regions and wine grape varieties, but this hypothesis requires further investigation. This research also represents the first vineyard study with U. oudemansii in a hot and dry climate, demonstrating its potential as an effective BCA under these conditions.

Three integrated programmes combining BCAs and natural products were also evaluated, namely BZ-AZ, BZ (CS + FC) and (CS + FC)-AZ, which resulted in significant and effective reductions of botrytis bunch rot control, equivalent to reductions achieved with C. sake alone. Thus, integrated programmes did not significantly improve botrytis bunch rot control in the conditions tested. However, a combination of strategies with multiple modes of action has been recommended to ensure more consistent disease control and overcome fluctuations in external factors (Guetsky et al., 2001; Elad & Stewart, 2004; Elmer & Reglinski, 2006).

Survival patterns for the two C. sake concentrations applied were similar and population numbers on flowers and the berry surface over time in 2009 and 2010 corresponded well with the findings reported by Cañamás et al. (2011). The combination of C. sake and Fungicover resulted in good field survival and persistence of this BCA on grape berries late in the growing season and occurred at a time when BCA persistence on the grape bunch may be variable (Holz & Volkmann, 2002). Populations of C. sake CPA-1 significantly declined across all treatments from 106 to 102 CFU g−1 host tissue between flowering and pre-bunch closure in both seasons, despite different meteorological conditions. During this period, many floral tissues abscise after the last flowering spray and there are also major physiological changes taking place as the pollinated flowers become berries. The volume of tissue on developing fruitlets increases, increasing the surface area of tissue not previously treated and thereby affecting the proportion of CFU recovered per gram. These two processes may explain the low population counts that were recovered prior to pre-bunch closure application.

Candida sake populations on berries after pre-bunch closure declined between each application and this pattern was repeated between the different C. sake treatments. The rate of decline was progressively lower in both seasons, providing some evidence for a gradual establishment of yeast populations on developing grape berries during the season, as originally hypothesized by Cañamás et al. (2011). It is also possible that survival was favoured by more moderate temperatures in the grape canopy in the final 21 days before harvest. Population declines between applications are dependent on external factors such as temperature, RH, UV radiation and rain exposure, while studies on the direct effects of these factors are currently being developed for a better understanding of C. sake population dynamics. Nonetheless, the survival pattern of C. sake CPA-1 was comparable to that of other antagonist yeasts that were field-applied: declining between spray applications, but, overall, remaining above 102 CFU g−1 (Schena et al., 2000). Some yeast population studies reported more stable populations over time in postharvest conditions (Schena et al., 1999; Zahavi et al., 2000), but botrytis bunch rot control was lower.

On BZ-(CS + FC) treated berries, C. sake populations declined after the first application at véraison, and this coincided with the driest period of the whole season, with high temperatures that may have decreased early establishment by C. sake. During this same period U. oudemansii populations were relatively stable. After the 21-days-preharvest spray of C. sake and Fungicover, C. sake populations reached predictable numbers, while U. oudemansii CFU counts decreased, initially suggesting some incompatibility between the early season BCA on this tissue and C. sake. However, previous preliminary laboratory studies and results from an identical experiment carried out in a conventional vineyard in the same region (data not shown) did not show any incompatibility between U. oudemansii and C. sake CPA-1. This research is the first report that quantifies the survival of the B. cinerea antagonist U. oudemansii on green grapevine tissues. Results indicated that populations of U. oudemansii were relatively stable between flowering and pre-bunch closure, an important attribute for a BCA dependent upon necrotic tissues that at the time of application are green. Ulocladium oudemansii CFU counts recovered from green flowers were similar to those reported by Elmer & Kohl (1998) on lily leaves treated with the B. cinerea antagonist U. atrum 385 (2 × 106 CFU mL−1). Low population counts on the grape berry surface indicated that this BCA found this ecological niche to be more hostile than the floral tissues. Nonetheless, the populations on the grape berries were relatively stable at 1 × 102 CFU mL−1 until 21 days before harvest.

Several fungi, particularly yeasts, can adversely affect grape fermentation and wine quality parameters. Consequently, microvinifications were carried out with treated grapes in both 2009 and 2010 to analyse possible effects of BCA treatments on fermentation. The progression of wine fermentation was similar among the different treatments and no relevant differences were detected in the wine quality parameters analysis in 2009 or 2010. Moreover, trained panel tasting did not reveal any fault in wines elaborated with C. sake-treated grapes in 2010.

All the biologically based strategies evaluated significantly reduced botrytis bunch rot incidence and severity, providing further evidence that alternatives to synthetic fungicides can achieve consistent, effective and practical control. Candida sake CPA-1, combined with Fungicover, is an attractive new BCA that is able to survive environmentally harsh conditions on the grape berry surface. Applications of this BCA resulted in high average efficacy for two seasons under a range of conditions in organic vineyards, indicating that this yeast-based BCA is a good candidate for future larger-scale field application with standard viticultural sprayers. Furthermore, wine quality was not adversely affected by C. sake applications and lower input treatments achieved promising results. These findings represent a significant advance towards commercialization of this BCA in the near future.


The authors are grateful to the University of Lleida for CCG's PhD grant and organic production research project, the European Union (INTERREG REDBIO project) for their financial support and the Catalan Government (for a travel-assistance grant) and the Plant and Food Research Institute for travel assistance for PAGE.