Scientia Horticulturae 165 (2014) 433–438

Contents lists available at ScienceDirect

Scientia Horticulturae

I

a

KIBPCBS

1

SrsiaLcrma

tPsa

Pv

0h

journa l h om epage: www.elsev ier .com/ locate /sc ihor t i

ntegrated control of green mold to reduce chemical treatment in post-harvest citrus fruits

r t i c l e i n f o

eywords:mazaliliological controlreventive applicationurative applicationacillus subtilisaccharomyces cerevisiae

a b s t r a c t

This study shows that Saccharomyces cerevisiae (ACB-K1 and ACB-CR1) and Bacillus subtilis (ACB-69 andACB-84) isolates perform differently on the control of green mold (Penicillium digitatum), depending onthe citrus variety. In ‘Murcott’ tangor, the yeast ACB-CR1 resulted in 47% of healthy fruits, which increasedto 67% when combined with imazalil 0.5 mL L−1. In the ‘Hamlin’ orange, ACB-CR1 (S. cerevisiae) provided87% control when applied alone. However, when combined with 0.5 and 1.0 mL L−1 of fungicide (thelowest doses), the efficiency of ACB-CR1 was decreased, yielding 76 and 78% healthy fruits, respectively.Both yeasts controlled green mold in the ‘Tahiti’ acid lime by 40% when used as a curative treatment;however, the ACB-K1 isolate that was applied as a preventive measure was the best antagonist, yielding73% healthy fruits. This yeast increased disease control, with healthy fruit percentages ranging from 84to 89% when the microorganism was combined with the lowest doses of imazalil. In general, B. subtilisisolates provided only slight disease control when tested in the three citrus fruit varieties during thisstudy. However, the results of preventive treatments with bacteria on ‘Tahiti’ acid lime fruits revealedan improvement in the degree of biocontrol. This study demonstrated the possibility of reducing theimazalil dose during the post-harvest citrus fruit treatment using a biocontrol agent without losing greenmold control efficiency under storage conditions (27 ◦C and 70% RH [relative humidity]). The preventiveapplication provided the best protection to ‘Tahiti’ acid lime fruits, suggesting that the mode of action of

these biocontrol agents is through competition, or even resistance induction, considering the specificityof the antagonist–host relationship within the context of pathogen control. The yeast isolates decreasedtheir antagonistic activity against P. digitatum under refrigeration (10 ◦C and 95% RH). However, the ACB-K1 isolate provided 100% disease control in ‘Tahiti’ acid lime fruits under these storage conditions whencombined with a quarter dose of imazalil, despite its low activity.

© 2013 Elsevier B.V. All rights reserved.

. Introduction

Green mold, which is caused by Penicillium digitatum (Pers.:Fr)acc., is one of the most significant diseases of post-harvest in cit-us fruits causing extensive losses during harvest, transport, andtorage processes (Eckert and Eaks, 1989). Synthetic fungicides,ncluding imazalil (IMZ) and thiabendazole (TBZ), have tradition-lly been used to control green mold (Ismail and Zhang, 2004;ahlali et al., 2005; Smilanick et al., 2006). However, the efficacy ofhemical treatments has often been limited in the face of pathogenesistance to these compounds, as well as concerns about environ-ent contamination and public health. Accordingly, a search for

lternative control strategies is needed (Pérez et al., 2011).The use of antagonistic microorganisms has emerged as a poten-

ial alternative to synthetic fungicides for controlling diseases.

revious research showed that the protective activity of the twoelected antagonist yeasts, Rhodosporidium kratochvilovae (LS11)nd Cryptococcus laurentii (LS28), is enhanced by combining them

Abbreviations: IMZ, imazalil; TBZ, thiabendazole; BCA, biological control agent;DA, potato dextrose agar; NYDA, nutrient yeast dextrose agar; ANOVA, analysis ofariance; CFU, colony forming unit.

304-4238/$ – see front matter © 2013 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.scienta.2013.11.019

with a low dosage of fungicides and/or natural adjuvants, and thatsuch strategies can control both resistant and sensitive strains offungal pathogens (Lima et al., 2006, 2011). Therefore, selected bio-control yeasts are very interesting candidates for their utilizationin integrated control strategies aimed at reducing the use of fungi-cides.

Among the biological control agents, Saccharomyces cerevisiaeand Bacillus subtilis have been used for the biocontrol of pathogensthat occur during the post-harvest period (Coelho et al., 2003;Leelasuphakul et al., 2008; Sharma et al., 2009; Kupper et al., 2013).The main mechanisms of action of yeasts during post-harvestdisease biocontrol include competition for space and nutrients, pro-duction of volatile metabolites, enzymes that degrade the plantpathogen cell wall (including �-1,3-glucanase and chitinase), hostresistance induction, mycoparasitism, and the ‘killer’ factor (a toxicpeptide) (Coelho et al., 2003; E.I-Tarabily and Sivasithamparam,2006). The Bacillus species have been promising in the controlof a variety of fungi that cause plant diseases, and their antago-nistic action against pathogens occurs through the production ofantibiotics (iturin, surfactin, and fengycin), by enzymes (chitinase,

�-1,3-glucanase) that degrade structural polymers of the fungaland by production of volatile antifungal agents (Leelasuphakulet al., 2006; Pinchuk et al., 2002). In certain situations, volatile

4 post-

ol(caad

bmtfitaa(tdtdttosm

itp

2

2

b(cPtCBacewsh

2

ddp28afsh

2

t

34 Integrated control of green mold to reduce chemical treatment in

rganic compounds that were secreted by B. subtilis cells wereinked to plant growth and plant systemic resistance inductionCompant et al., 2005). The efficiency of a biological control agentan be enhanced with the addition of chemical control. Errampallind Brubacher (2006) demonstrated the control of P. expansum inpples by integrating the use of Pseudomonas syringae with cypro-inil, which provided the best results.

Research on post-harvest disease has focused on application ofiocontrol agents after harvest. However, application after harvestay be too late for the biocontrol agents to effectively compete with

he decay pathogens already established on or in fruit tissues in theeld. Antagonistic microorganisms that have curative action, con-rolling pre-existing infections, that prevents subsequent infectionsnd retards fungal sporulation are desirable. On the other hand, thebility of the treatments to protect the fruit from future infectionspreventive activity) should be evaluated. In commercial situationshe reinfection of the same fruit or on healthy fruits may occururing handling and processing within the packing house, wherehe surface wounds can be infected by the pathogen for severalays. The application of products based on curative or preven-ive forms of antagonistic microorganisms and an understanding ofheir modes of action against plant pathogens may benefit the fieldf biocontrol. Usall et al. (2008) reported that the combination ofodium carbonate with Pantoea agglomerans (CPA-2) bacteria wasore efficient against P. digitatum.Considering all these factors, this study was aimed at evaluat-

ng the use of S. cerevisiae and B. subtilis to control P. digitatum,he causal agent of citrus green mold, with or without a chemicalroduct (imazalil).

. Material and methods

.1. Cultivars

The cultivars in this study were ‘Murcott’ tangor, a hybridetween tangerine (Citrus reticulata Blanco) and sweet orangeCitrus sinensis [L.] Osb.). The fruits were harvested from the Agri-ultural Experimental Station at the Sylvio Moreira APTA (Sãoaulo Agribusiness Technology Agency) Citrus Production Cen-er/IAC (Agronomic Institute of Campinas) (Estac ão Experimental,entro APTA Citros Sylvio Moreira/IAC), Cordeirópolis, São Paulo,razil. A ‘Hamlin’ orange tree (C. sinensis (L.) Osb.) Cv. and ‘Tahiti’cid lime (Citrus latifolia Tanaka) were purchased from commer-ial orchards in the state of São Paulo (Brazil). The fruits in thesexperiments were not subjected to regular commercial treatment,hich is usually applied post-harvest, and they were used on the

ame day or stored up to 2 weeks (5 ± 1 ◦C and 95% RH [relativeumidity]) prior to their use in various assays.

.2. Pathogen

A highly virulent strain of P. digitatum (PF-1) was obtained fromecayed oranges and used to artificially inoculate the fruit. Coni-ial suspensions for fruit inoculation were obtained as follows: theathogen was grown on potato dextrose agar (PDA) for 7 days at7 ◦C. Ten microliters of sterile distilled water with 0.01% Tween0 was dispensed into Petri dishes. Conidia were scraped from thegar using a sterile loop. The suspension was subsequently trans-erred to a test tube and was sonicated for 5 min to facilitate conidialuspension, and the concentration was adjusted with the aid of aemocytometer.

.3. Antagonists

B. subtilis strains ACB-84 and ACB-69 were obtained fromhe APTA Center Citros ‘Sylvio Moreira’ IAC, Cordeirópolis, São

harvest citrus fruits / Scientia Horticulturae 165 (2014) 433–438

Paulo, Brazil. The S. cerevisiae strains (ACB-CR1 and ACB-K1) wereobtained from the Laboratory for Biochemistry and Plant Pathol-ogy at the University of São Paulo (ESALQ), Piracicaba, São Paulo,Brazil. These strains were selected by assay in vitro and in vivo ofantagonistic action of the BCAs against P. digitatum (Kupper et al.,2013).

The activated culture was maintained on NYDA (nutrient yeastdextrose agar) medium at 27 ◦C for 48 h and transferred, and thecell suspension was used as inoculum for mass production in afermentation system at 27 ◦C in the dark.

2.4. Preparation and fermentation of biocontrol agents

B. subtilis was grown in the presence of foliar fertilizer underglutamic fermentation of molasses to 0.5% with agitation for 72 h(Bettiol et al., 2005). The residue, which is known as Ajifol®, wasused to grow the bacteria because it contains carbon sources, nitro-gen, and salts in addition to being low-cost and in common use bycitrus orchards. A liquid medium containing potato dextrose wasused to produce S. cerevisiae with stirring for 72 h.

2.5. In vitro sensitivity of biocontrol agents to imazalil

The aim of this experiment was to evaluate B. subtilis (ACB-69 and ACB-84) and S. cerevisiae (ACB-CR1 and ACB-K1) for theirsensitivity to the imazalil (500 g L−1, active ingredient) fungicidethat is used to control green mold with the intention of developingintegrated control methods.

B. subtilis and S. cerevisiae isolates were grown on NYDA (nutri-ent yeast dextrose agar) with stirring at 250 rpm at 30 ◦C for 48 h.A 200 �L aliquot of the microorganism suspension (108 CFU mL−1)was subsequently plated on Petri dishes with NYDA medium using aDrigalski handle. The cells from each biological control agent (BCA)were subsequently scattered evenly over the culture medium. InPetri dishes containing the BCAs, two sterilized discs of filter paper(12.7 mm diameter) were placed and were soaked in the fungicideimazalil (IMZ) at concentrations of 0.5; 1.0; 2.0 (as recommendedby the manufacturer); 4.0 and 8.0 mL L−1, and five plates were madeby treatment.

The cultures were incubated at 27 ◦C for 72 h. Five replicateswere used per treatment. An evaluation was performed on the basisof the presence or absence of an inhibition halo between the filterpaper discs and the bacterial or yeast growth.

2.6. Use of biocontrol agents and imazalil in the integratedcontrol of citrus fruit green mold

Citrus fruits from the ‘Murcott’ tangor, ‘Hamlin’ orange, and‘Tahiti’ acid lime cultivars with no postharvest treatment werewashed using a soft sponge, neutral detergent, and water and weresurface-disinfected with 0.7% (v/v) sodium hypochlorite for 3 min.The fruits were wounded at two equidistant points on the equato-rial region with a sterilized stylus at a depth of 3 mm and inoculatedwith 20 �L of P. digitatum conidial suspension (1 × 105 spores mL−1)24 h prior to starting the different treatments. A preventive biolog-ical control effect was only done for ‘Tahiti’ acid lime fruits underambient conditions; therefore, those fruits were treated 24 h priorto inoculation with the plant pathogen.

The treatments were as follows: (i) each BCA was applied sep-arately, with cell suspensions corresponding to 1 × 108 CFU mL−1

for 2 min; (ii) imazalil 0.5 mL L−1 + fermented broth from each BCA

(1 × 108 CFU mL−1) for 2 min; (iii) imazalil 1.0 mL L−1 + fermentedbroth from each BCA (1 × 108 CFU mL−1) for 2 min; (iv) imazalil2.0 mL L−1 (as recommended by the manufacturer); (v) the control,in which fruits were inoculated with P. digitatum (1 × 105 spores

Integrated control of green mold to reduce chemical treatment in post-h

012345678910

8.04.02.01.00.50

ACB-CR1 ACB-K1

Saccharomyces cerevisiae

concentration of IMZ (mL L-1)

nº

of

inh

ibit

ion

halo

012345678910

8.04.02.01.00.50

ACB-84 ACB-69

nº

of

inh

ibit

ion

halo

Bacillus sub�lis

concentration of IM Z (mL L-1)

a a a

bc ab ab

a

d

cd

a

b b

a

bc

c c c

Fs

mi

d1dtt

2

aubtT

3

3

aitot0w

3c

3

c

ig. 1. Sensitivity of Saccharomyces cerevisiae (ACB-K1 and ACB-CR1) and Bacillusubtilis (ACB-69 and ACB-84) isolates to various doses of imazalil.

L−1) and treated with water only); (vi) imazalil 0.5 mL L−1; (vii)mazalil 1.0 mL L−1.

Following inoculation and treatment, fruits were stored underifferent conditions, cooling conditions (10 ± 1 ◦C and 95% RH) for5 days and at 27 ◦C and 70% RH (ambient conditions) for sevenays. During the ‘Hamlin’ orange fruits test, the mean air tempera-ure was 20 ◦C (70 ± 5% RH), and the evaluation was performed onhe 15th day after the beginning of the test.

.7. Statistical analysis

A completely randomized design was used with three replicates,nd each replicate consisted of 15 fruits for each cultivar. An eval-ation was performed to determine disease incidence, as assessedy the percentage of healthy fruits, and the data were submittedo an analysis of variance (ANOVA) and comparison of means usingukey’s test at 5% probability.

. Results

.1. In vitro sensitivity of biocontrol agents to imazalil

Testing ‘in vitro’ of the sensitivity of strains bacterial (ACB-69nd ACB-84) and yeast (ACB-K1 and ACB-CR1) at different doses ofmazalil under laboratory conditions is shown in Fig. 1. In general,he highest fungicide doses 4.0 and 8.0 mL L−1 affected the growthf S. cerevisiae, and yet the ACB-K1 isolate had the highest sensitivityo fungicide, with inhibition halos being present at doses above.5 mL L−1. Bacterial isolates were the most sensitive to imazalil,ith growth inhibition being observed in all assessed doses.

.2. Use of biocontrol agents and imazalil in the integratedontrol of citrus fruit green mold

.2.1. ‘Murcott’ TangorDisease control was 63 and 53% in treatments only with the S.

erevisiae ACB-K1 and ACB-CR1 isolates, respectively. ACB-K1 and

arvest citrus fruits / Scientia Horticulturae 165 (2014) 433–438 435

ACB-CR1 treatment resulted in 77 and 67% healthy fruits, respec-tively, combined with the lowest dose of fungicide (0.5 mL L−1).These combinations efficiently inhibited disease onset comparedto the fungicide treatment at the recommended dose (2.0 mL L−1)with no significant differences between them (Fig. 2A).

There was a slight increase in disease incidence, with 53%healthy fruits being observed, regardless of the isolate type, whenyeasts were combined with a half dose of imazalil (1.0 mL L−1).This result may be partly explained by data found in vitro (Fig. 1),in which the ACB-K1 yeasts were sensitive to fungicide at doseshigher than 0.5 mL L−1. Following the combination of B. subtilis iso-lates with imazalil to control green mold in ‘Murcott’ tangor fruits,a combination of ACB-69 and fungicide at a dose of 0.5 mL L−1

resulted in 73% disease control with no significant difference fromthe imazalil treatment that was applied at the recommended dose.Conversely, bacterial isolate ACB-84 combined with the same doseof fungicide did not lead to satisfactory disease control (27% healthyfruits). ‘Murcott’ tangor fruits that were treated only with ACB-69and ACB-84 bacterial isolates had 0 and 47% healthy fruits, respec-tively.

3.2.2. ‘Hamlin’ orangeFruits treated with doses of 0.5, 1.0, and 2.0 mL L−1 imazalil had

the highest numbers of healthy fruits (87, 97.8, and 100%, respec-tively) (Fig. 2B).

We found that ACB-CR1 yeast promoted more antagonisticactivity against P. digitatum, with 87% control, in contrast to theACB-K1 isolate, which yielded only 13% healthy fruits. Whenthe yeasts were combined with imazalil, ACB-CR1 plus 0.5 and1.0 mL L−1 resulted in 76 and 78% disease control, respectively; andACB-K1 yielded 56 and 91% asymptomatic fruits with the samefungicide concentrations.

Bacterial isolates alone showed no potential to control greenmold in ‘Hamlin’ orange fruits; the percentages of healthy fruitsranged from 71 to 93% (with half the dose of imazalil) and 47 to 71%(with a quarter dose), which occurred only when both were com-bined with different doses of imazalil, and ACB-84 always providedthe highest protection values in the mixture.

3.2.3. ‘Tahiti’ acid limeThe ‘Tahiti’ acid lime fruits were effectively protected with

imazalil at a rate of 0.5 mL L−1 when stored under ambient con-ditions (27 ◦C and 70% RH), yielding 93% healthy fruits, regardlessof antagonist addition (Fig. 2C). The percentages of healthy citrusfruits ranged from 93 to 98% with no difference between the fungi-cide dose that was recommended by the manufacturer (2.0 mL L−1),and when S. cerevisiae isolates were combined with the lowestdoses of fungicide (0.5 and 1.0 mL L−1).

With respect to the B. subtilis isolates, the ACB-69 yielded42% healthy fruits when combined with a half dose of imazalil(1.0 mL L−1) or with a quarter dose of imazalil (0.5 mL L−1). Whenthe same isolate was applied without fungicide, it yielded 33.33%asymptomatic fruits (Fig. 2C). The ACB-84 isolate (without fungi-cide) had no efficacy when compared with ACB-69, in combinationswith the different doses of fungicide (0.5 and 1.0 mL L−1) whichassessed 42 and 27% of healthy citrus fruits, respectively.

Fruit treatments with different doses of imazalil resulted in thehighest percentages of healthy fruits (80 to 95%) when applied to‘Tahiti’ acid lime fruits and in storage under cold room conditions(10 ◦C and 95% RH). Under these storage conditions, the yeast iso-lates decreased their antagonistic activity against P. digitatum; thatis, fruits treated only with ACB-K1 and ACB-CR1 had 0 and 13%

healthy fruits, respectively (Fig. 2D).

When combined with a quarter dose of imazalil under cold stor-age, the ACB-K1 isolate treatment yielded 100% control in ‘Tahiti’acid lime fruits, despite its low activity, whereas the fungicide alone

436 Integrated control of green mold to reduce chemical treatment in post-

Fig. 2. Incidence of healthy citrus fruit (%) after inoculation with Penicilliumdigitatum (1.0 × 106 conidia mL−1), treatment with biological control agents (BCAs)(1 × 108 CFU mL−1) alone or in combination with doses of imazalil (0.5, 1.0 and2.0 mL L−1); (A) tangor ‘Murcott’ fruit incubation for 7 days at 27 ◦C and 70% RH; (B)

harvest citrus fruits / Scientia Horticulturae 165 (2014) 433–438

at the same dose had 20% diseased fruits, while the percentage ofhealthy fruits found at the dose recommended by the manufacturer(2.0 mL L−1) was 91% (Fig. 2D).

B. subtilis isolates yielded a relatively significant pathogencontrol under refrigerated conditions (10 ◦C and 95% RH) whencombined with the reduced dose of imazalil (Fig. 2D). The ACB-69isolate combined with 0.5 mL L−1 imazalil resulted in 85% diseasecontrol, and fruits treated with the combination of ACB-69 and1.0 mL L−1 imazalil or only ACB-69 provided 30 and 2.22% asymp-tomatic fruits, respectively. The ACB-84 isolate without fungicidewas the least efficient in disease control, and when combined withdoses of 0.5 and 1.0 mL L−1 imazalil, the percentage of healthy fruitsreached 75 and 60%, respectively.

The percentage of healthy fruits were 84 and 89% when theS. cerevisiae ACB-K1 isolate was combined with the lowest dosesof fungicide (0.5 and 1.0 mL L−1 imazalil). It is critical to empha-size that the mixture of ACB-K1 yeast with a half dose of imazalil(1.0 mL L−1) inhibited disease onset as effectively as treatment withthe recommended fungicide dose (2.0 mL L−1) for disease control,with no significant difference being detected between them. Thistreatment yielded 73% asymptomatic fruits when the ACB-K1 iso-late was applied preventively with no fungicide. The ACB-CR1isolate was not efficient for controlling disease.

With respect to the B. subtilis isolates, the ACB-84 yielded 40, 67,and 80% healthy fruits when applied alone or in combination with0.5 and 1.0 mL L−1 imazalil, respectively. The ACB-69 isolate with-out fungicide and in combinations with doses of 0.5 and 1.0 mL L−1

fungicide yielded 53, 58, and 73% fruits without green mold symp-toms, respectively.

4. Discussion

The present study shows that S. cerevisiae (ACB-K1 and ACB-CR1)and B. subtilis (ACB-69 and ACB-84) isolates have different levels ofantagonistic activity, depending on the variety of citrus fruit, andalso storage conditions.

In the case of the ‘Murcott’ tangor, the ACB-CR1 yeast wasthe biocontrol agent that best controlled the disease, the controlefficiency increased when the antagonist was combined with aquarter dose imazalil (0.5 mL L−1). S. cerevisiae isolate ACB-CR1 con-trolled the disease better when applied by itself to ‘Hamlin’ orange.However, when ACB-CR1 was combined with the lowest doses offungicide, 0.5 and 1.0 mL L−1 the control efficiency decreased.

In the case of the ‘Tahiti’ acid lime, the ACB-K1 isolate was thebest antagonist when applied as a preventative agent, increase offruits without disease symptoms, when combined with the lowestimazalil doses (0.5 and 1.0 mL L−1), this microorganism increasedcontrol efficiency (Fig. 2E).

In general, B. subtilis (ACB-69 and ACB-84) isolates provided lit-tle disease control when tested in the citrus fruit varieties, despitehaving demonstrated its efficacy against P. digitatum in in vitrotests (Kupper et al., 2013). However, the results of preventive treat-ments with bacteria in ‘Tahiti’ acid lime fruits indicated improveddisease control. As shown in Fig. 2E, ACB-84 and ACB-69 yieldedhealthy fruits when applied alone or when combined with 0.5 and1.0 mL L−1 imazalil, decrease in the quantity fruits without green

mold symptoms, comparing preventive and curative treatment.

This study demonstrated the possibility of reducing the imazalildose during the post-harvest treatment of citrus fruits using bio-control agents. This green mold control can be accomplished

orange ‘Hamlin’ fruit incubation for 7 days at 20 ◦C and 70% RH; (C) acid lime ‘Tahiti’fruit incubation for 7 days at 27 ◦C and 70% RH; (D) acid lime ‘Tahiti’ fruit (curativetreatment) incubation for 15 days at 10 ◦C and 95% RH; (E) acid lime ‘Tahiti’ fruit(preventive treatment) incubation for 7 days at 27 ◦C and 70% RH.

post-h

w(hcdAtiihedd(2

(cfrttceccSrsauoi1tseeaf(

mbfnfwtaeea(efs

tBAeh

au

Integrated control of green mold to reduce chemical treatment in

ithout the loss of efficiency under ambient storage conditions27 ◦C and 70% RH). The ACB-K1 yeast yielded 77, 93, and 100%ealthy fruits in the ‘Murcott’ tangor and ‘Tahiti’ acid lime whenombined with a quarter of treatment of imazalil and preventiveoses, respectively. The same dose of fungicide combined with theCB-CR1 isolate yielded 76% healthy fruits in the ‘Hamlin’ orange

ree. The combination of a biocontrol agent with reduced doses ofmazalil will help to reduce the chemical residue in citrus fruitsn addition to reducing the amount of this product in the packingouse. With respect to the bacterial isolates, ACB-69 promoted dis-ase control in ‘Tahiti’ acid lime fruits when combined with a halfose of imazalil. The combination of biocontrol agents with loweroses of fungicides has already been reported by several authorsKinay et al., 2001; Papadopoulou-Mourkidou, 1991; Usall et al.,001; Zhou et al., 2002).

This same hypothesis was successfully confirmed by Lima et al.2006 and 2011) integration studies of biocontrol yeasts and fungi-ides thiabendazole (TBZ), boscalid (BOSC), cyprodinil (CYPR), andenhexamid (FENH) in the integrated control of sensitive andesistant isolates of Botrytis cinerea and P. expansum within the pro-ections of stored apples. Despite the fact that research has shownhat integration of BCA with low doses of fungicide can enhanceontrol of postharvest fungal rots (Ippolito et al., 2004), Lima andmployees (2006 and 2011), demonstrated that such integrationan also manage fungicide resistant isolates of B. cinerea more effi-iently. The results are in full agreement with Chand-Goyal andpotts (1997); they observed an improved control of postharvestots on stored apples caused by either isolates of P. expansum sen-itive or resistant to benzimidazoles by using biocontrol yeasts and

reduced rate of TBZ, whereas the fungicide applied alone yieldednsatisfactory control. In fact, in the presence of a resistant isolatef B. cinerea, alone or combined with a sensitive isolate, TBZ wasneffective since it acts by a specific mechanism of action (Delp,995), while the activity of the biocontrol yeast isolate C. lauren-ii LS28, is based on different mechanisms such as competition forpace and nutrients, hyperparasitism with the production of lyticnzymes and resistance to oxidative stress in fruit wounds (Castoriat al., 2001, 2003). These mechanisms are complex and not specific,ffecting multiple sites in the pathogen; in addition, it is unlikelyor a fungal pathogen to develop resistance to a biocontrol yeastLima et al., 2006).

The curative application of these antagonists as a disease treat-ent does not favor biocontrol because the mode of action of the

iological control agents most likely occurs through by competitionor space and nutrients, which could explain the reduced effective-ess of green mold control in the ‘Murcott’ tangor (0 to 63% healthy

ruits) and ‘Tahiti’ acid lime (0 to 40%) fruits. Therefore, P. digitatumas presumably already established at the site of infection at the

ime of biocontrol agent application, resulting in disadvantages forntagonist development. This hypothesis was also reported by Usallt al. (2008) in P. digitatum biocontrol tests with Pantoea agglom-rans (CPA-2), which only enabled limited healing activity of thentagonist when applied alone. In previous study, Teixidó et al.2001) showed that the P. agglomerans bacterial isolate was able toffectively colonize the wound at the citrus fruit peel, thereby inter-ering with and preventing the pathogen infection, despite havinghown limited growth.

In our study, the preventive application yielded the best pro-ection of ‘Tahiti’ acid lime citrus fruits, suspecting that both the. subtilis (ACB -69 and ACB -84) and S. cerevisiae (ACB -CR1 andCB -K1) modes of action may be accomplished by competition orven the induction of resistance, considering the specificity of the

ost-antagonist relationship in relation to pathogen control.

Tests conducted with acid lime fruits allowed us not only tossess the mode of biocontrol agent application but also to eval-ate the performance of fruits under different storage conditions,

arvest citrus fruits / Scientia Horticulturae 165 (2014) 433–438 437

after treatment with antagonists and in combination or not with themost common fungicide for disease control under typical packing-house conditions in Brazil.

Under refrigerated conditions (10 ◦C and 95% RH) the yeast iso-lates decreased their antagonistic activity against P. digitatum; thatis, the fruits treated with ACB-K1 and ACB-CR1 (Fig. 2D). How-ever, the ACB-K1 isolate yielded 100% disease control in ‘Tahiti’ acidlime fruits when combined with a quarter dose of imazalil underthese fruit storage conditions, despite its low activity, whereasthe fungicide alone at the same dose resulted in the reduction ofhealthy fruits (Fig. 2D) at the dose recommended by the manu-facturer (2.0 mL L−1). In contrast to the present study, Kinay et al.(2001) found that combination of the antagonist and imazalil orthiabendazole fungicides (200 �g mL−1) inhibited disease onsetas effectively as the 1000 �g mL−1 fungicide concentration whentesting the A15/1 yeast isolate for the control of green and bluemolds in grapefruit under refrigeration (10 ± 1 ◦C). The combina-tion of different control methods favored disease control becausethe pathogen has its infection and disease onset activity reducedunder low temperature conditions.

Considering that discrepancies in biocontrol efficacy betweencontrolled laboratory conditions and commercial circumstancescan occur and have various causes, the aim of this study was toprovide the optimal conditions for pathogen development, includ-ing the concentration of inoculum used, wounding and favorabletemperature and humidity conditions, especially when studieswere used to assess the effectiveness of biocontrol agents whencombined with low doses of fungicide during citrus fruit processingunder ambient conditions.

Above all, this study demonstrated that it is possible to manip-ulate ambient conditions and day-to-day practices, which, in thiscase, refers to the packing house, to favor the biocontrol agent andto reduce the use of fungicides, thereby protecting not only theenvironment but also the consumers of citrus fruits. Further investi-gations are needed to test this integrated strategy under large-scaleconditions.

Acknowledgements

The financial support of the Fapesp (The State of São PauloResearch Foundation) (Proceeding no. 2007/03692-0), whichprovided a scholarship to Cristiane Moretto (Proceeding no.2007/02045-0), is gratefully acknowledged.

References

Bettiol, W., Kupper, K.C., Goes, A., De Moretto, C., Corrêa, É.B., 2005. Mass productionof Bacillus subtilis and Trichoderma viride for the control of Phyllosticta citricarpa(teleomorph: Guignardia citricarpa). Summa Phytopathol. 31 (3), 276–278.

Castoria, R., De Curtis, F., Lima, G., Caputo, L., Pacifico, S., De Cicco, V., 2001. Aureoba-sidium pullulans (LS30) an antagonist of postharvest pathogens of fruit: studyon its modes of action. Postharvest Biol. Technol. 22, 7–17.

Castoria, R., Caputo, L., De Curtis, F., De Cicco, V., 2003. Resistance of postharvestbiocontrol yeasts to oxidative stress: a possible new mechanism of action. Phy-topathology 93, 564–572.

Chand-Goyal, T., Spotts, R.A., 1997. Biological control of postharvest diseases of appleand pear under semi-commercial conditions using three saprophytic yeasts.Biol. Control 10, 199–206.

Coelho, A.R., Hoffmann, F.L., Hirooka, E.Y., 2003. Biocontrol by yeasts: perspectivesin aplication for fruits safety: a review. Semin. Ciênc. Agrár. 24, 337–358.

Compant, S., Duffy, B., Nowak, J., Clement, C., Barka, E., 2005. Use of plant growth-promoting bacteria for biocontrol of plant diseases: principles mechanisms ofaction, and future prospects. Appl. Environ. Microbiol. 71, 4951–4959.

Delp, C.J., 1995. Benzimidazole and related fungicides. In: Lyr, H. (Ed.), ModernSelective Fungicides. VEB Gustav Fischer Verlag, Longman Group, UK Ltd., Jena,London, pp. 291–303 (Chapter 14).

Eckert, J.W., Eaks, I.L., 1989. Postharvest disorders and diseases of citrus fruits. In:

Reuter, W., Calavan, E.C., Carman, G.E. (Eds.), The Citrus Industry. Univ. CaliforniaPress, Berkeley, pp. 179–260.

E.I-Tarabily, K.A., Sivasithamparam, K., 2006. Potential of yeasts as biocontrol agentsof soil-borne fungal plant pathogens and as plant growth promoters. Myco-science 47, 25–35.

4 post-

E

I

I

K

K

L

L

L

L

L

P

P

P

S

38 Integrated control of green mold to reduce chemical treatment in

rrampalli, D., Brubacher, N.R., 2006. Biological and integrated control of posthar-vest blue mold (Penicillium expansum) of apples by Pseudomonas syringae andcyprodinil. Biol. Control 36, 49–56.

ppolito, A., Nigro, F., Schena, L., 2004. Control of postharvest diseases of fresh fruitsand vegetables by preharvest application of antagonistic microorganisms. In:Dris, R., Niskanen, R., Jain, S.M. (Eds.), Crop Management and Postharvest Hand-ling of Horticultural Products. Disease and Disorders of Fruits and Vegetables,vol. 4. Science Publisher Inc., Enfield, NH, USA, pp. 1–30, 333.

smail, M., Zhang, J., 2004. Post-harvest citrus diseases and their control. OutlooksPest Manag. 15, 29–35.

inay, P., Yildiz, M., Yildiz, F., Delen, N., Tosun, N., 2001. Control of postharvest Peni-cillium decay of citrus fruits with antagonistic yeasts and chemical fungicides.Acta Hortic. 553, 383–388.

upper, K.C., Cervantes, A.L.L., Klein, M.N., da Silva, A.C., 2013. Evaluation of antago-nistic microorganisms. Saccharomyces cerevisiae and Bacillus subtilis for controlof Penicillium digitatum. Rev. Bras. Frutic. 35, 425–436.

ahlali, R., Serrhini, M.N., Jijakli, M.H., 2005. Development of a biological controlmethod against postharvest diseases of citrus fruits. Commun. Agric. Appl. Biol.Sci. 70, 47–58.

eelasuphakul, W., Sivanunsakul, P., Phongpaichit, S., 2006. Purification, character-ization and synergistic activity of �-1,3-glucanase and antibiotic extract froman antagonistic Bacillus subtilis NSRS 89-24 against rice blast and sheath blight.Enzym. Microb. Technol. 38, 990–997.

eelasuphakul, W., Hemmanee, P., Chuenchitt, S., 2008. Growth inhibitory prop-erties of Bacillus subtilis strains and their metabolites against the green moldpathogen (Penicillium digitatum Sacc.) of citrus fruit. Postharvest Biol. Technol.48, 113–121.

ima, G., De Curtis, F., Piedimonte, D., Spina, A.M., De Cicco, V., 2006. Integrationof biocontrol yeast and thiabendazole protects stored apples from fungicidesensitive and resistant isolates of Botrytis cinerea. Postharvest Biol. Technol. 40,301–307.

ima, G., Castoria, R., De Curtis, F., Raiola, A., Ritieni, A., De Cicco, V., 2011. Integratedcontrol of blue mould using new fungicides and biocontrol yeasts lowers lev-els of fungicide residues and patulin contamination in apples. Postharvest Biol.Technol. 60, 164–172.

apadopoulou-Mourkidou, E., 1991. Postharvest-applied agrochemicals and theirresidues in fresh fruits and vegetables. J. Assoc. Off. Anal. Chem. 74, 743–765.

érez, E., Blanco, O., Berreta, C., Dol, I., Lado, J., 2011. Imazalil concentration for in vitromonitoring of imazalil resistant isolates of Penicillium digitatum in citrus packinghouses. Postharvest Biol. Technol. 60, 258–262.

inchuk, I.V., Bressollier, P., Sorokulova, I.B., Verneuil, B., Urdaci, M.C., 2002. Ami-

coumacin antibiotic production and genetic diversity of Bacillus subtilis strainsisolated from different habitats. Res. Microbiol. 153, 269–276.

harma, R.R., Singh, D., Singh, R., 2009. Biological control of postharvest diseasesof fruits and vegetables by microbial antagonists: a review. Biol. Control 50,205–221.

harvest citrus fruits / Scientia Horticulturae 165 (2014) 433–438

Smilanick, J.L., Brown, G.E., Eckert, J.W., 2006. Postharvest citrus diseases and theircontrol. In: Wardowski, W.F., Miller, W.M., Hall, D.J., Grierson, W. (Eds.), FreshCitrus Fruit. , second ed. Florida Science Source, Inc., Longboat Key, FL, pp.339–396.

Teixidó, N., Usall, J., Palou, L., Asensio, A., Nunes, C., Vinas, I., 2001. Improving controlof green and blue molds of oranges by combining Pantoea agglomerans (CPA-2)and sodium bicarbonate. Eur. J. Plant Pathol. 107, 658–694.

Usall, J., Teixidó, N., Torres, R., Ochoa de Eribe, X., Vinas, I., 2001. Pilot tests of Can-dida sake (CPA-1) applications to control postharvest blue mold on apple fruit.Postharvest Biol. Technol. 21, 147–156.

Usall, J., Smilanick, J., Palou, L., Denis-Arrue, N., Teixidó, N., Torres, R., Vinas, I., 2008.Preventive and curative activity of combined treatments of sodium carbonatesand Pantoea agglomerans CPA-2 to control postharvest green mold of citrus fruit.Postharvest Biol. Technol. 50, 1–7.

Zhou, T., Northover, J., Schneider, K.E., Lu, X., 2002. Interactions between Pseu-domonas syringae MA-4 and cyprodinil in the control of blue mold and graymold of apples. Can. J. Plant Pathol. 24, 154–161.

Cristiane Moretto a

Antonio Lucas Lima Cervantes b

Antonio Batista Filho a

Katia Cristina Kupper b,∗a Instituto Biológico, Avenida Conselheiro Rodrigues

Alves, São Paulo, SP, 1.252-CEP 04014-002, Brazilb Centro de Citricultura Sylvio Moreira, Instituto

Agronômico, Rod. Anhanguera, Km 158,Cordeirópolis, SP, CEP 13490-970 Brazil

∗ Corresponding author. Tel.: +55 19 3546 1399;fax: +55 19 3546 1399.

E-mail address: katia@centrodecitricultura.br(K.C. Kupper)

8 August 2013

21 October 2013

13 November 2013Available online 15 December 2013