Trop. plant pathol.

DOI 10.1007/s40858-015-0008-x

Efficacy of Yeast in the Biocontrol of Bacterial Fruit Blotch

in Melon Plants
Edilaine Alves de Melo & Rosa de Lima R. Mariano &
Delson Laranjeira & Liliana Andrea dos Santos &
Luciana de Omena Gusmão & Elineide Barbosa de Souza

Received: 15 May 2014 / Accepted: 3 November 2014

# Sociedade Brasileira de Fitopatologia 2015

Abstract Bacterial fruit blotch (BFB), caused by Acidovorax 1 Introduction

citrulli, is the most important bacterial disease affecting melon
plants in Brazil. The aim of the present study was to analyze Bacterial fruit blotch (BFB), caused by Acidovorax citrulli
the efficacy of yeasts for the biological control of BFB, ap- (Schaad et al.) Schaad et al., is one of the most severe diseases
plied by spraying to protect seedlings, foliage and fruits, and affecting melons (Cucumis melo L.). This disease is a major
as a seed treatment. Additionally, the in vitro activity of the problem in the northeast region of Brazil, which is the most
yeasts and plant growth promotion were evaluated. Among 60 important area for the production of this fruit (IBGE 2013).
strains, LMA1 (Rhodotorula aurantiaca), CC-2 (Pichia Since 1997, when the disease was first detected in Brazil
anomala) and LMS (Rhodotorula glutinis) were the most ef- (Assis et al. 1999), BFB has occurred with high severity in
ficient for seedling protection. Among the three selected melon crops, frequently causing high losses (Oliveira et al.
strains, only LMA1 and CC-2 maintained efficacy when test- 2003). When infected, the fruit becomes unsuitable for sale,
ed on foliage and seeds, with reductions in the disease index exhibiting small oily spots that expand and become brown and
and area under the disease progress curve of 58.6 and 47.2 %, necrotic with or without ruptures in the center. Internally dry
respectively. Furthermore, LMA1 and LMS protected fruits. rot occurs due to the bacterial colonization of the fruit pulp.
In vitro, neither of the strains inhibited bacterial growth, pro- Acidovorax citrulli can affect several organs of the melon
duced killer toxins, or showed competition for nutrients with plant during different stages of development. Therefore, strat-
the pathogen. None of the three yeasts promoted the growth of egies for controlling BFB should include the treatment of
melon plants. Therefore, these yeasts have potential for BFB seeds and foliage, and the use of bacteria-free seeds is a pri-
biocontrol in different disease stages and may be used in the mary measure that should be employed. Physical treatments of
integrated management of BFP. the seeds and chemical sprays of the crop (Latin and Hopkins
1995) have been recommended, although they do not control
the disease efficiently. Selection for BFB-resistant plants has
Keywords Biological control . Killer toxins . Plant growth been performed in accessions and varieties of melon and
promotion other cucurbitaceous plants (Hopkins and Thompson
2002; Bahar et al. 2009; Carvalho et al. 2013). However,
inconsistent results have been obtained. As chemical and
physical treatments to control the disease are not effective
Section Editor: Wagner Bettiol
and resistant cultivars of melon are not available, other strat-
egies are being investigated.
E. A. de Melo : R. de Lima R. Mariano : D. Laranjeira :
L. A. dos Santos : L. de Omena Gusmão
Antagonistic bacteria have been tested for the biocon-
Departamento de Agronomia, Universidade Federal Rural de trol of BFB in melon and watermelon. The treatment of
Pernambuco, 52171-900 Recife, Pernambuco, Brazil watermelon seeds with Acidovorax avenae subsp. avenae
Manns AAA99-2 suppressed the disease and reduced BFB
E. Barbosa de Souza (*)
transmission by the seeds (Fessehaie and Walcott 2005).
Departamento de Biologia, Universidade Federal Rural de
Pernambuco, 52171-900, Recife, Pernambuco, Brazil Furthermore, the application of AAA99-2 and Pseudomonas
e-mail: [email protected] fluorescens Migula A506 to flowers suppressed the epiphytic
 Trop. plant pathol.

growth of the pathogen and reduced infestation in the seeds of (40 g dextrose, 10 g neopeptone and 17 g agar/L) sup-
fruits produced from the treated flowers (Fessehaie and plemented with yeast extract (1.5 g/L) and chloramphenicol
Walcott 2005). Bacillus Cohn and Paenibacillus Ash et al. (50 mg/L). Plates were incubated at 28 °C for 48 h. The yeast
strains, obtained endophytically or epiphytically from melon colonies were separated based on appearance and color, and
or other cultures, have also shown potential in the biocontrol transferred to glass tubes containing SDA medium. Thirty-
of BFB through both seed microbiolization and seedling spray seven yeast strains isolated from melon fruits acquired from
(Oliveira et al. 2006; Medeiros et al. 2009). However, in a farmer’s markets belonging to the Culture Collection of the
review of the literature, only two papers with yeasts were Laboratório de fungos do solo (UFRPE) were also used. Yeast
found. Pichia anomala (Hansen) Kurtzman sprayed on leaves suspensions were prepared from cultures in SDA or SD de-
of Hami melon (C. melo var. saccharinus Naud) was effective pending on the experiment performed. The concentration of
in reducing the incidence and severity of BFB. Seeds treated suspensions was adjusted to 1.5×108 cells/mL using the
with a metabolic extract of P. anomala resulted in a reduction McFarland scale.
of disease incidence in plants (Wang et al. 2009). Furthermore, The three most efficient strains were identified through
Rhodotorula aurantiaca (Saito) Lodder LMA1, Rhodotorula molecular analysis using primers specific for the region
glutinis (Fresen) Harrison LMS and P. anomala CC-2, effec- D1/D2 26S rDNA, and the ITS1 and ITS4 primers of the ITS
tive in the control of BFB, did not present additive or syner- region.
gistic effects when combined with silicon (Conceição et al.
2014). Even though yeasts have the unique feature of high 2.2 Protection of Melon Seedlings with Yeasts
competition for space and nutrients, they are also susceptible
to low humidity and are not easily formulated, which could Melon seeds from the cultivar Mandacaru F1 were sown in
explain the low number of tests with yeasts. 500 mL plastic pots containing a soil:humus (1:1v/v) mixture.
The studies of biocontrol with yeasts may prove successful Plants were inoculated with yeasts at 6 days after emergence.
in Brazilian conditions with strains adapted to main melon- Ten milliliters of yeast suspensions were applied until runoff
producing regions. Therefore, the aim of the present study was using a DeVilbiss atomizer. After 24 h, the cotyledonar leaves
to evaluate the efficacy of yeasts in the biological control of were inoculated until runoff with an A. citrulli Aac1.12 sus-
BFB through spraying to protect seedlings, foliage and fruits, pension (Araújo et al. 2005). Immediately after inoculation,
and through seed treatment. Additionally, in vitro activities of the plants were transferred to a growth chamber (temperature
yeasts against A. citrulli and melon growth promotion were of 32.5±2 °C and relative humidity of 53.0±2 %) for 24 h.
analyzed. After inoculation, the plants were assessed every day from
6 days and severity of the disease was determined using a
descriptive scale (Araújo et al. 2005), with scores ranging
2 Materials and Methods from 0 to 5 (0=asymptomatic seedlings; 1–4=asymptomatic
hypocotyls and symptoms on one or both cotyledons, with
2.1 Acquisition, Cultivation and Identification of Strains lesions corresponding to 1–25, 26–50, 51–75 and 76–
100 %, respectively; and 5=total necrosis of the cotyledons
A strain of A. citrulli Aac1.12 [Culture Collection of the and hypocotyls, regardless of damping-off). The following
Laboratório de Fitobacteriologia, Universidade Federal Rural parameters were determined from the data obtained: the incu-
de Pernambuco (UFRPE), Recife, PE, Brazil] was grown in bation period (IP), the disease index (DI) calculated using the
nutritive-yeast-dextrose agar (NYDA) (10 g dextrose, 3 g beef McKinney (1923) formula, and the area under the disease
extract, 5 g yeast extract, 3 g peptone and 18 g agar/L) at 28 °C progress curve (AUDPC) (Shaner and Finney 1977). The
for 36 h. To prepare the inoculum, the bacteria were diluted in plants were maintained in a greenhouse at 32.5±2 °C and
sterile distilled water (SDW), and the suspension was adjusted 53.0±2 % relative air humidity measured using a digital
to 3.4×107 CFU/mL using a spectrophotometer (Analyzer thermo-hygrometer (Incoterm Ltda.). The experimental de-
500 M, Analyzer). sign was completely randomized with five replicates, and each
Yeasts were isolated from symptomless melon plants replicate consisted of five seedlings. The strains with the best
grown in the crop areas of Petrolina and Mossoró, in the states performance were selected to be evaluated for the protection
of Pernambuco and Rio Grande do Norte, respectively. The of leaves and fruits, and seed treatments, as well as the growth
leaves were cut into small fragments, and 1 g samples were promotion assessment and in vitro studies of biocontrol
placed in glass tubes with 9 mL SDW containing 50 mg/L of mechanisms.
chloramphenicol. After incubation in an ultrasonic bath In all experiments, acibenzolar-S-methyl (ASM) (Bion 500
(ULTRAsonik 5B, Denstply Neytech) for 15 min, the suspen- WG, Syngenta) (50 mg/L) and distilled water were used as
sion was diluted and aliquots of 100 μL were pipetted onto control. The chemical surfactant Tween 80 was added to treat-
9 mm Petri dishes containing Sabouraud-dextrose-agar (SDA) ments at 0.03 % (v/v).
Trop. plant pathol.

2.3 Protection of Melon Foliage with Yeasts 2.5 Melon Seed Treatment with Yeasts

Melon plants were cultivated in 500 mL plastic pots contain- Melon seeds were disinfected by immersion in sodium hypo-
ing a soil:humus (1:1v/v) mixture. After 15 days, the first pair chlorite (0.5 % of active chlorine) for 10 min, rinsed in SDW
of true leaves was individually sprayed until runoff with R. and air-dried for 24 h. Subsequently, the seeds were immersed
aurantiaca LMA1, R. glutinis CC-2 and P. anomala LMS in 250 mL of A. citrulli Aac1.12 suspension for 30 min and
yeasts suspensions. Twenty-four hours later, the leaves were dried at room temperature (25±2 °C) for 16 h (Wang et al.
sprayed until runoff with an A. citrulli Aac1.12 suspension 2009). This seeds were immersed in 100 mL of R. aurantiaca
(Silveira et al. 2003). The plants were placed in a humid LMA1, R. glutinis CC-2 and P. anomala LMS yeast suspen-
chamber at pre- and post-inoculation for 24 h. After pathogen sions for 30 min. The inoculated and treated seeds were dried
inoculation, the plants were evaluated daily for 10 days at room temperature (±27 °C) for 16 h. The seeds were planted
for the appearance of symptoms and the severity of on expanded polystyrene trays containing a soil:humus (1:1
the disease. The assessment was performed using a de- v/v) mixture. After emergence, the trays were placed in a hu-
scriptive scale (Silveira et al. 2003), with scores ranging mid chamber for 24 h. The seedlings were monitored daily for
from 0 to 6 (0=asymptomatic leaves; 1–6=1–5 %, 6– 6 days for the appearance of symptoms and severity of the
12 %, 13–37 %, 38–62 %, 63–87 % and 88–100 % disease. The disease severity was estimated using a descriptive
infected foliar area, respectively). The IP, DI and scale (Araújo et al. 2005) with scores ranging from 0 to 5 (0=
AUDPC were calculated from the data obtained as described seedlings without symptoms; 1=seedlings with marginal le-
above. The plants were maintained in a greenhouse (29.3± sions up to 50 % on one or both cotyledons; 2=seedlings with
2 °C and 60.0±2 % relative air humidity). The experimental marginal lesions up to 75 % on both cotyledons, a few lesions
design was completely randomized with ten replicates, each in the center of the blade and slight leaf deformation; 3=seed-
replicate with eight plants each, and two leaves per plant. lings with marginal lesions of up to 100 % on both cotyledons,
several lesions in the center of the blade, acute leaf deforma-
tion and stunting; 4=seedlings with marginal lesions of up to
2.4 Protection of Melon Fruits with Yeasts 100 % on both cotyledons, several lesions in the center of the
blade progressing to the hypocotyl, total leaf deformation and
Mature fruits of melon cultivar Mandacaru F1 (average stunting; 5=total necrosis on the cotyledons and hypocotyl,
diameter and brix, 43 cm and 10°, respectively) were wilting and death). The IP, DI and AUDPC were calculated
inoculated using the subepidermal injection method from the data obtained. The plants were maintained in a green-
(Somodi et al. 1991). After washing with soap and water, house (32.2±2 °C and 49.5±2 % relative air humidity). The
surface sterilization with sodium hypochlorite (0.5 % a.i.) for experimental design was completely randomized with ten rep-
5 min, and washing with distilled water and drying, the fruits licates, and each replicate consisted of 30 seedlings.
were sprayed with yeast strains R. aurantiaca LMA1, R.
glutinis CC-2 and P. anomala LMS. After 24 h, each 2.6 In vitro Sensitivity of A. citrulli to Yeasts
fruit was marked with three parallel lines along its
length. One hundred microliters of bacterial suspension The antibiosis of the strains R. aurantiaca LMA1, R. glutinis
(3.4×107 CFU/mL) was then injected with a hypodermic CC-2 and P. anomala LMS against A. citrulli Aac1.12 was
syringe just beneath the rind surface, in the intercellular tested using the modified methodology of Wang et al. (2009).
spaces, at three points on each line. Fruits were maintained An aliquot of 500 μL of bacterial suspension (3.4×107 CFU/
in a greenhouse at 28.3±2 °C and 45.5±2 % relative air hu- mL) was seeded onto NYDA medium in 15 cm Petri dishes.
midity. The experimental design was completely randomized, Using a sterilized 6 mm cork borer, nine wells spaced at 3-cm
with six replicates (fruits), and each replicate consisted of nine intervals were made in the culture medium. Aliquots of
inoculation points. The fruits were evaluated 8 days after in- 100 μL of liquid yeast cultures were placed in each well and
oculation for BFB symptoms by using a descriptive scale with SDW was used as a control. Plates were incubated at 28±2 °C
scores ranging from 1 to 6 (1=without rind darkening, light for 48 h, and the inhibition of bacterial growth was recorded.
pulp discoloration and without cavity formation in the pulp; The experimental design was completely randomized with
2=with rind darkening, light pulp discoloration and without five replicates, and each replicate consisted of one Petri dish.
cavity formation in the pulp; 3=with rind and pulp darkening To evaluate the production of antibacterial metabolites by
and without cavity formation in the pulp; 4=without rind the three yeast strains, aliquots of 60 μL of each yeast suspen-
darkening, light pulp discoloration and with cavity for- sion (1.5×108 cells/mL) were individually placed in 150 mL
mation in the pulp; 5=with rind darkening, light pulp Erlenmeyer flasks containing 60 mL of liquid SD medium,
discoloration and cavity formation in the pulp; and 6=with which were incubated at 28 °C for 4 days and shaken every
rind and pulp darkening and cavity formation in the pulp). 24 h. The yeast suspensions were subsequently centrifuged at
 Trop. plant pathol.

13,000 g for 15 min, and the supernatant was sterilized The shoots and roots were separately weighed on an analytical
through filtration. The cell-free culture filtrate was tested for scale (Mark 210A, Tecnal) to calculate the fresh bio-
the inhibition of A. citrulli as described above for liquid yeast mass of shoots and roots. Subsequently, the plant mate-
cultures with the same experimental design. rial was placed inside paper bags, oven-dried at 45 °C for
The killer activity in the R. aurantiaca LMA1, R. glutinis 5 days and weighed again to calculate the dry biomass of the
CC-2 and P. anomala LMS yeast strains was analyzed in 9- shoots and roots.
mm dishes containing SDA medium buffered with citrate-
phosphate at different pH values (4.5, 5.5, 6.5, 7.5 and 8.5). 2.9 Statistical Analyses
The yeast culture was streaked onto one of the edges of the
plate, and four horizontal streaks of the pathogen were placed All assays were replicated to determine the consistency of the
close and perpendicular to the yeast streak. Plates were incu- results. Given that significant differences in the variances of
bated at 28 °C for 48 h, and the bacterial growth inhibition was the experimental replicates were not observed (P≤0.05), the
assessed. The experimental design was completely random- data were evaluated as replicates in time. The assumptions of
ized with five replicates, and each replicate consisted of one the analysis of variance (ANOVA) were verified using the
Petri dish. Shapiro-Wilk and Levene tests in the software Statistix ver-
sion 9.0 (Florida State University). The means were compared
2.7 Competition for Nutrients between A. citrulli and Yeasts by the Scott-Knott test (P≤0.01) using the software SAEG
version 9.0 (Universidade Federal de Viçosa) or the LSD
The competition for nutrients between strains of A. citrulli and Test using the software Statistix version 9.0.
the biocontrol yeasts (R. aurantiaca LMA1, R. glutinis CC-2
and P. anomala LMS) was evaluated with the Biolog Gen III
system (Biolog Inc.), based on the utilization of 71 carbon 3 Results
sources. Inoculum suspensions were prepared in the inoculat-
ing fluid IF-A, after adjusting the transmittance to 98 %, from 3.1 Isolation and Identification of Yeasts
cultures that had been grown in Biolog Universal Growth
(BUG), medium at 31 °C for 36 h. After, 100 μl of the sus- Twenty-three strains were obtained from the crop areas of
pension were distributed per well into Biolog microplates, Petrolina and Mossoró and 37 from the Culture Collection
which were incubated at 33 °C for 22 to 36 h. The presence of the Laboratory of Soil Fungi at UFRPE. Among the 60
of microbial growth in the wells was determined by the purple strains evaluated (Table 1), the LMA1, CC-2 and LMS yeasts
color that indicated TZC [2,3,5-triphenyl tetrazolium chloride] strains were identified as R. aurantiaca, R. glutinis and
reduction. The determination of the number of carbon sources P. anomala, respectively, based on molecular features.
used in common by the antagonist and the pathogen was cal-
culated by the equation, INO=NCSCPA/TNCSP, where INO 3.2 Protection of Melon Seedlings with Yeasts
was the index of niche overlap; NCSCPA was the number of
carbon sources used in common by the pathogen and antago- From the 60 yeasts tested, R. aurantiaca LMA1, R. glutinis
nist; and TNCSP was the total number of carbon sources used CC-2 and P. anomala LMS were the most efficient. They
by the pathogen. It was established that if INO>0.9 both mi- significantly (P ≤0.01) increased the IP (6.7, 6.6 and 5.0 days,
croorganisms are competent to occupy the same niche but if respectively) and reduced the DI (1.2, 2.0 and 14.0 %, respec-
INO<0.9 they both occupied different niches (Halfeld-Vieira tively) and AUDPC (0.2, 0.3 and 1.8, respectively) in relation
et al. 2015). to the control (IP 2.0 days, DI 64.0 % and AUPDC 9.6)
(Table 1). LMA1 and CC-2 did not differ from the ASM-
2.8 Effect of Yeast on Melon Growth acibenzolar-S-methyl.

Melon plants were grown in a greenhouse in 500 mL plastic 3.3 Protection of Melon Foliage and Fruits with Yeasts
pots containing a soil: humus (1:1v/v) mixture for 15 days,
and the leaves subsequently were sprayed until run off with The R. aurantiaca LMA1, R. glutinis CC-2 and P. anomala
R. aurantiaca LMA1, R. glutinis CC-2 and P. anomala LMS LMS yeasts were efficient in the protection of melon foliage
suspensions. The plants were maintained in a greenhouse at against BFB when sprayed on the leaves 24 h before inocula-
32.2±2 °C and 49.5±2 % relative air humidity. The experi- tion with A. citrulli. They significantly reduced (P≤0.05) the
mental design was completely randomized with ten replicates, DI (44.2, 39.7 and 43.3 %, respectively) and AUDPC (13.1,
and each replicate consisted of ten plants. Fifteen days after 14.0 and 14.7, respectively) compared with the control (DI
treatment, the plants were removed from the pots, and their 95.8 % and AUDCP 24.8). Only treatments with ASM and
roots were rinsed in distilled water to remove soil fragments. LMS significantly increased (P≤0.05) the IP (Table 2).
Trop. plant pathol.

Table 1 Effect of yeast strains in

the incubation period (IP, in days), Treatmenta IP DI AUDPC Treatment IP DI AUDPC
disease index (DI, in %) and area
under the disease progress curve Control 2.0 eb 64.0 a 9.6 a LMA-3 2.8 e 38.0 c 5.6 c
(AUDPC) of bacterial fruit blotch CA-4 2.3 e 35.6 c 5.6 c CC-8 2.8 e 42.0 c 6.0 c
(Acidovorax citrulli) in melon CPS-3 2.3 e 38.4 c 6.7 b CC-17 2.9 e 30.4 d 4.6 c
seedlings, in the greenhouse
CPS-9 2.3 e 48.8 b 7.3 b CC-3 2.9 e 44.4 b 5.1 c
LMA-2 2.4 e 45.2 b 7.2 b LMA-18 2.9 e 41.6 c 6.0 c
LMA-15 2.4 e 36.8 c 5.9 c CC-14 2.9 e 30.4 d 4.6 c
CA-7 2.4 e 37.6 c 5.9 c CA-3 2.9 e 31.2 d 4.9 c
LMA-4 2.4 e 47.6 b 7.3 b LMA-7 3.0 d 42.4 c 6.3 c
CC-5 2.4 e 39.2 c 5.9 c CPS-6 3.1 d 36.4 c 4.9 c
CC-24 2.4 e 36.8 c 5.5 c LMA-11 3.1 d 34.0 c 4.7 c
LMA-14 2.4 e 40.8 c 6.2 c CC-10 3.2 d 30.0 d 4.6 c
LMA-5 2.5 e 44.8 b 7.3 b CA-5 3.3 d 41.6 c 5.7 c
CA-1 2.5 e 41.6 c 6.3 c LMA-19 3.3 d 35.6 c 5.4 c
CA-6 2.5 e 50.0 b 7.5 b CC-16 3.4 d 29.2 d 3.6 d
a
CA, CPS, CC and SPS, Yeast LMA-17 2.5 e 38.8 c 6.2 c CC-12 3.4 d 27.6 d 3.7 d
isolated from melon fruits
acquired at farmer’s markets, LMA-12 2.5 e 42.8 c 7.1 b LMA-10 3.4 d 26.0 d 3.7 d
belonging to the Culture CC-26 2.6 e 35.6 c 5.2 c CPS-4 3.6 d 23.2 d 2.9 d
Collection of the Laboratório de CA-2 2.6 e 44.8 b 6.7 b CA-9 3.6 d 17.6 d 2.8 d
fungos do solo (UFRPE); LMA,
LMA-13 2.6 e 45.6 b 6.2 c CC-4 3.8 c 24.8 d 2.8 d
LMS and LMC, yeast isolated
from melon leaves from the crop CA-12 2.6 e 37.2 c 6.1 c CA-11 3.9 c 28.8 d 4.4 d
areas of Petrolina (PE) and CA-10 2.6 e 41.2 c 6.0 c LMA-8 3.7 c 21.6 d 3.6 d
Mossoró (RN) CPS-8 2.6 e 46.8 b 6.3 c CC-6 4.1 c 20.4 d 2.8 d
b
Mean of ten replicates with five LMA-6 2.7 e 42.0 c 6.3 c LMA-20 4.1 c 27.2 d 4.1 d
plants each. Each mean represents
data from two experiments that LMC 2.7 e 33.6 c 5.1 c CC-9 4.2 c 30.0 d 4.8 c
did not differ from each other. CPS-2 2.7 e 39.6 c 6.0 c CC-25 4.4 c 28.8 d 3.1 d
Within the column, means CPS-5 2.7 e 37.2 c 5.3 c CC-15 4.4 c 21.2 d 3.0 d
followed by the same letter are
SPS-3 2.7 e 35.2 c 5.2 c LMA-21 4.7 b 24.8 d 3.8 d
not significantly different
(P≤0.01) by the Scott-Knott test LMA-9 2.7 e 38.4 c 5.6 c LMS* 5.0 b 14.0 e 1.8 e
*LMS, Rhodotorula glutinis; CC- LMA-16 2.7 e 47.0 b 6.3 c CC-2* 6.6 a 2.0 f 0.3 f
2, Pichia anomala; LMA1, CC-1 2.8 e 34.0 c 4.0 d ASM* 6.7 a 2.0 f 0.2 f
Rhodotorula aurantiaca; ASM, CA-8 2.8 e 27.2 d 4.8 c LMA-1* 6.7 a 1.2 f 0.2 f
acibenzolar-S-methyl

The yeasts LMA1 and LMS were efficient in protecting treatments with yeasts were less effective than those
melon fruits by significantly reducing (P≤0.05) the disease with ASM for BFB control. None of these three yeasts
severity (1.5 and 1.4 respectively) in relation to the control reduced seed disease transmission, as the disease incidence
(3.0) (Table 2). CC-2 did not significantly differ (P≤0.05) was 100 % in all treatments.
from the control and also from the other yeast strains.

3.4 Melon Seeds Treatment with Yeasts 3.5 In vitro Sensitivity of A. citrulli to Yeasts

Melon seedlings grown from seeds treated with R. aurantiaca None of the yeast strains R. aurantiaca LMA1, R. glutinis CC-
LMA1 and R. glutinis CC-2 exhibited a lower severity of BFB 2 and P. anomala LMS inhibited A. citrulli Aac1.12 growth in
symptoms with reductions in DI (4.3 and 47.8 %, re- NYDA medium. A clear area around the yeast suspension
spectively) and AUDPC (3.8 and 4.8, respectively) in wells was not observed after 48 h of incubation at 28 °C.
relation to the control (DI 65.6 % and AUDPC 7.0), Pathogen growth inhibition using cell-free yeast culture fil-
but only LMA1 and ASM increased the IP (Table 2). trate was also not observed. In addition, the three strains were
The LMS isolate was not effective in melon seed treat- negative for the production of killer toxins against A. citrulli in
ment. Notably, in the plant protection and seed assays, SDA medium at the various pH values tested.
 Trop. plant pathol.

Table 2 Effect of Rhodotorula aurantiaca LMA1, Rhodotorula (AUDPC) of bacterial fruit blotch (Acidovorax citrulli) spraying on
glutinis CC-2 and Pichia anomala LMS in the incubation period (IP, in foliage and fruit and seed treatment of melon, in the greenhouse
days), disease index (DI, in %) and area under the disease progress curve

Treatmenta Foliage Fruit Seed

Incubation period Disease index AUDPC Severity Incubation period Disease index AUDPC

Control 3.0 cb, c 95.8 ad 24.8 a 3.0 ae 1.1 bf 65. 6 ad 7.0 a

LMA1 3.2 bc 44.2 b 13.1 b 1.5 b 2.4 a 43. 1 b 3.8 d
ASM 4.4 a 23.5 c 7.7 c - 2.7 a 28. 3 c 2.8 e
CC-2 3.2 bc 39.7 b 14.0 b 2.3 ab 1.3 b 47. 8 b 4.8 c
LMS 3.4 b 43.3 b 14.7 b 1.4 b 1.3 b 61. 7 a 6.2 b
CV (%) 10.7 4.8 17.1 37.6 17.7 21. 4 18.8
a
ASM, Acibenzolar-S-methyl
b
Mean value represents the data obtained from two experiments that did not differ in their variances. Within the column, mean values followed by the
same letter are not significantly different (P≤0.05) according to the LSD test
c
Mean of 10 replicates with eight plants each
d
Transformed data (Log x+1)
e
Mean of 12 replicates (fruits) with nine points of inoculations each
f
Mean of 20 replicates with 30 seeds each

3.6 Competition for Nutrients between A. citrulli and Yeasts Colletotrichum spp., Fusarium spp., Penicillium spp.,
Rhizopus spp. and Sclerotinia sclerotiorum (Lib.) de Bary
There was no competition for nutrients between A. citrulli and (Haissam 2011).
R. aurantiaca LMA1, R. glutinis CC-2 and P. anomala LMS. Among the 60 yeasts sprayed onto melon seedlings,
The INO indexes (0.11, 0.19 and 0.31, respectively) were the strains R. aurantiaca LMA1, R. glutinis CC-2 and
below 0.9. P. anomala LMS provided protection against BFB (Table 1).
These results are significant because melon plants show high
3.7 Effect of Yeasts on Melon Plant Growth susceptibility to BFB during their initial and final develop-
mental stages, i.e., seedlings and fruits (Bahar et al. 2009).
The R. aurantiaca LMA1, R. glutinis CC-2 and P. anomala Moreover, control in seedlings is important because under
LMS did not promote (P≤0.05) melon plant growth compared the conditions of transplant houses, with dense plant popula-
with the control, as assessed through the variables fresh bio- tions, high temperatures and humidity, and overhead irriga-
mass of shoots, dry biomass of shoots, fresh biomass of roots tion, A. citrulli is easily disseminated among seedlings
and dry biomass of roots, thus exhibiting behavior similar to (Burdman and Walcott 2012) and BFB incidence is increased.
ASM. These strains did not show adverse effects, except for
LMS, which reduced the dry biomass of shoots (1.1 g) in Table 3 Effect of yeasts on melon plant growth
relation to control (1.5 g) (Table 3).
Treatment Fresh Dry Fresh Dry
biomass of biomass of biomass of biomass of
shoots (g) shoots (g) roots (g) roots (g)

4 Discussion Control 11.6 aba 1.5 ab 2.0 a 0.6 a

Acibenzolar-S-metil 14.3 a 1.2 bc 2.5 a 0.6 a
Studies on the use of yeasts for controlling bacterial diseases (ASM)
Rhodotorula 12.5 ab 1.8 a 3.3 a 0.8 a
are scarce, particularly for diseases affecting shoots. However, aurantiaca LMA1
different genera of yeasts are efficient control agents for post- Pichia anomala CC-2 14.7 a 1.7 a 2.4 a 0.7 a
harvest fungal pathogens diseases in fruits, vegetables and Rhodotorula glutinis 9.2 b 1.1 c 1.8 a 0.6 a
cereals. YieldPlus (Cryptococcus albidus (Saito) Skinner), LMS
Shemer (Metschnikowia fructicola Kurtzman & Droby) a
Means of 20 replicates with 10 plants each. Each mean represents data
and Nexy (Candida oleophila Montrocher isolate O) are from two experiments that did not differ from each other. Within the
recommended for the treatment of post-harvest diseases column, means followed by the same letter are not significantly different
caused by Aspergillus niger Tiegh, Botrytis cinerea Pers., (P≤0.05) by Kruskal-Wallis test
Trop. plant pathol.

The protection of melon seedlings against A. citrulli infection Species of the genera Pichia and Rhodotorula have
was also achieved through spraying with Paenibacillus been identified as potential agents in plant disease control,
lentimorbus (Dutky) Pettersson, Rippere, Yousten and Priest particularly for postharvest infections. P. anomala is an
MEN2, which reduced DI (81 %), AUDPC (88 %) and inci- efficient agent for the biological control of fungi of agricul-
dence (77 %) (Medeiros et al. 2009). The selection of biocon- tural importance (Walker 2011), such as B. cinerea and
trol agents for seedlings has many advantages, such as low Penicillium spp. in apple (Haissam 2011) and other fungi
space and time requirements, and can be easily performed in a and Enterobacteriaceae present in cereal grains (Olstorpe
greenhouse. et al. 2012). The use of R. glutinis as an efficient biological
In the BFB cycle, leaf lesions are important sources of control agent in postharvest diseases of apples (Qin et al.
inoculum for the infection of the fruits (Latin and Hopkins 2003), pears (Zhang et al. 2008) and other fruits, and the use
1995). The yeasts strains R. aurantiaca LMA1, R. glutinis of Rhodotorula sp. (Rh1) in the control of soft rot in Chinese
CC-2 and P. anomala LMS were also efficient in BFB control cabbage (Mello et al. 2011) have been previously reported.
when sprayed on leaves, resulting in DI and AUDPC reduc- Acibenzolar-S-methyl (ASM) was selected as a positive
tions of up to 58.6 and 47.2 %, respectively, indicating that control because it is registered for use against BFB in melons
these yeasts can be applied in the field to reduce the amount of in Brazil (AGROFIT 2003) and has demonstrated efficacy in
inoculum on leaves available to be spread to the fruits. Fruits controlling the disease compared with other inductors and
are highly susceptible to BFB (Bahar et al. 2009), and under pesticides (Cabral et al. 2010; Conceição et al. 2014; Sales
optimal conditions of temperature and humidity, a total loss Júnior et al. 2007). However, since ASM is toxicological class
might occur in affected fields because symptomatic fruits can- III (AGROFIT 2003), new strategies must be found for BFB
not be sold (Latin and Hopkins 1995). Although plant protec- management emphasizing the importance of biocontrol with
tion provided less disease control than seedling protection, the yeast. Moreover, according to Conceição et al. (2014) the
overall levels of protection obtained were significant com- spraying of ASM in foliage resulted in lower levels of BFB
pared with other studies. The strain P. anomala 0732-1, control than were provided by yeast LMA1 and LMA1 plus
sprayed on Hami melon leaves, reduced the DI of BFB by silicon.
82.3 % (Wang et al. 2009); the Rhodotorula sp. LD-19 re- The mechanisms involved in the control of BFB by
duced the severity of tomato soft rot by 21.2 % (Gomes et al. R. aurantiaca LMA1, R. glutinis CC-2 and P. anomala LMS
2005); and three strains of Cryptococcus suppressed the pop- were not completely explained by the results of our study.
ulation of Erwinia amylovora (Burrill) Winslow by 65 % in These yeasts did not inhibit the growth of A. citrulli, either
the detached stigmas of apple flowers (Pusey et al. 2009). in liquid culture or cell-free culture filtrate, i.e., these yeasts
A. citrulli can be introduced into a particular area through did not act through antibiosis or the production of killer toxins.
infected seeds (Assis et al. 1999), and planting healthy seeds is The production of molecules with direct activity against path-
the primary measure recommended for BFB control (Latin and ogens, called antibiosis, is an important mechanism used by
Hopkins 1995). Moreover, the treatment of seeds with antag- yeast that is beneficial to plants. Nevertheless, Wang et al.
onists reinforces this control (Medeiros et al. 2009), as infected (2009) found that out of 463 yeast strains obtained from the
or contaminated seeds produce infected seedlings and spread leaves and flowers of cucurbits, 24 yeasts exhibited antibiosis
bacteria, resulting in increased disease incidence (Latin and against A. citrulli, and the cell-free culture filtrate of isolate
Hopkins 1995). In seed lots containing one seed infected with 0732-1 inhibited bacterial growth. The production of killer
different concentrations of bacteria, the transmission levels of toxins is a general feature of yeast strains from different gen-
A. citrulli ranged from 16.7 to 100 % (Dutta et al. 2011). In the era, including Saccharomyces (Hansen) Meyen, Hansenula,
seed treatment, the yeasts LMA1 and CC-2 confirmed their Kluyveromyces Van der Walt, Pichia and several others
efficacy in the control of BFB, however they did not reduce (Platania et al. 2012). However, the ability to produce killer
seed disease transmission (Table 2). Wang et al. (2009) also toxins is not an inherent feature of these species and varies
observed a significant reduction in BFB severity when Hami among the strains (Coelho et al. 2011). P. anomala produces
melon seeds infected with A. citrulli were treated with cell killer toxins against a broad spectrum of fungi, yeasts, bacteria
filtrate of P. anomala 0732-1. However, these results were and viruses (Walker 2011), and exhibits antibacterial activity
not different from those obtained using chemical treatments. against Erwinia (Polonelli and Morace 1986). The R. glutinis
Rhodotorula aurantiaca LMA1 was the only yeast isolate isolate 166 has also been described as killer yeast (Coelho et al.
efficient in the protection of harvested fruits against infection 2011). In relation to the production of enzymes associated with
by A. citrulli (Table 2). Similarly, for postharvest control, induction of resistance, increases in the activity of polyphenol
nine yeasts reduced the severity of Fusarium rot in mel- oxidase after foliar spraying with silicon or LMA1, and
on fruits from 10.4 to 70.9 % (Cruz 2010). These results increases of ascorbate peroxidase activity after foliar spraying
show the potential of these fungi for the biocontrol of different with LMA1 and LMA1 plus silicon were observed by
diseases of melon. Conceição et al. (2014) for the control of BFB. In addition,
 Trop. plant pathol.

nutrient competition was also not detected because each one Acknowledgments The authors thank the Fundação de Amparo à
Ciência e Tecnologia do Estado de Pernambuco (FACEPE) for the schol-
the three yeasts did not utilize a high number of the nutrients
arship awarded to EAM and the Conselho Nacional de Desenvolvimento
used by A. citrulli. Considering that none of these mechanisms Científico e Tecnológico (CNPq) for the research fellowship to RLRM
of action were found, it is reasonable to propose that induction and EBS. We would also like to thank Dr. Rejane Pereira Neves from the
of resistance might be involved in this biocontrol mechanism, Departamento de Micologia of the Universidade Federal de Pernambuco
(UFPE) for identifying the yeast strains.
as previously observed by Lu et al. (2013).
Yeasts have also been associated with the promotion of
plant growth (Rosa et al. 2010). The major direct actions of
microorganisms in promoting plant growth are associated References
with the production of phytohormones and assistance in plant
nutrition (biological nitrogen fixation and solubilization of AGROFIT (2003) Sistema de Agrotóxicos Fitossanitários 2003.
phosphates and other nutrients) (Vassilev et al. 2006). In the Available at: http://extranet.agricultura.gov.br/agrofit_cons/
principal_agrofit_cons. Accessed on 26 Jan 2015
present study, R. aurantiaca LMA1, R. glutinis CC-2 and Araújo DV, Mariano RLR, Michereff SJ (2005) Métodos de inoculação
P. anomala LMS did not promote the growth of melon plants, de Acidovorax avenae subsp. citrulli em melão. Summa Phytopathol
but they, for the most part, did not have deleterious effects 31:69–73
(Table 3). To promote the growth of sugar beet plants, Assis SMP, Mariano RLR, Silva-Hanlin DMW, Duarte V (1999)
Mancha-aquosa do melão causada por Acidovorax avenae subsp.
Shalaby and El-Nady (2008) used S. cerevisiae yeast for the
citrulli no Estado do Rio Grande do Norte. Fitopatol Bras 24:191
immersion of seeds; spray application on leaves and direct Bahar O, Kritzman G, Burdman S (2009) Bacterial fruit blotch of melon:
application to the soil. These authors observed that the latter screens for disease tolerance and role of seed transmission in path-
technique was the most adequate, resulting in an increase of ogenicity. Eur J Plant Pathol 123:71–83
leaf area, length and diameter of roots, and fresh and dry root Burdman S, Walcott R (2012) Acidovorax citrulli: generating basic and
applied knowledge to tackle a global threat to the cucurbit industry.
weight. In recent studies, Torulaspora globosa (Klocker) Van
Mol Plant Pathol 13:805–815
der Walt & Johannsen isolate 1S112 did not provide a signif- Cabral CP, Gama MAS, Alexandre ER, Mariano RLR, Silveira EB
icant increase in the height or dry matter of sorghum shoots, (2010) Efeito de acibenzolar-S-metil, mananoligossacarídeo e
although this yeast exhibited a capacity for indole acetic acid bioflavonóides cítricos no controle da mancha-aquosa e no
production in vitro (Rosa et al. 2010). These authors proposed crescimento do meloeiro. Trop Plant Pathol 35:119–123
Carvalho FCQ, Santos LA, Dias RCS, Mariano RLR, Souza EB (2013)
that the application of the agent through spraying on the leaves
Selection of watermelon genotypes for resistance to bacterial fruit
did not reach the soil and root system, which are the sites blotch. Euphytica 190:169–180
where the yeast would act in promoting plant growth. This Coelho AR, Nóbrega GMA, Pagnocca FC, Hoffmann FL, Harada K,
same outcome was observed in the present study, as the yeast Hirooka EY (2011) Avaliação do potencial antagônico de leveduras,
suspensions were also applied through spraying on the leaves. visando biocontrole de deterioração por Penicillium expansum.
Ciências Agrarias 32:1879–1892
The ASM inductor also did not influence plant growth, in Conceição CS, Felix KCS, Mariano RLR, Medeiros EV, Souza EB
contrast with the results of Cabral et al. (2010), who detected a (2014) Combined effect of yeast and silicon on the control of bac-
physiological cost through reductions in height and fresh bio- terial fruit blotch in melon. Sci Hortic 174:164–170
mass of melon shoots of 24.5 and 41.4 %, respectively, due to Cruz TM (2010) Potencial de leveduras no controle biológico da
this inductor. Only P. anomala LMS (Table 3) strain showed podridão-de-Fusário em frutos de meloeiro. PhD Thesis,
Universidade Federal Rural de Pernambuco. Recife, PE, Brazil
an adverse effect through a reduction in the dry biomass of the Dutta B, Scherm H, Gitaitis RD, Walcott RR (2011) Acidovorax citrulli
shoots of melon plants. This result might be attributed to the seed inoculum load affects seedling transmission and spread of bac-
cost of transferring metabolites involved in growth for the terial fruit blotch of watermelon under greenhouse conditions. Plant
synthesis of compounds related to plant defense. Dis 96:705–711
Fessehaie A, Walcott RR (2005) Biological control protect watermelon
BFB has been detected in many developmental stages of
blossoms and seed from infection by Acidovorax avenae subsp.
cucurbits. Therefore, the selection of biological control agents citrulli. Phytopathology 95:413–419
of the pathogen should consider this characteristic, i.e., differ- Gomes AMA, Silveira EB, Mariano RLR (2005) Tratamento pós-colheita
ent disease stages, to provide results that are more reliable in com cálcio e microrganismos para controle da podridão-mole em
conditions of natural occurrence. The yeasts R. aurantiaca tomate. Hortic Bras 23:108–111
Haissam JM (2011) Pichia anomala in biocontrol for apples: 20 years of
(LMA1), P. anomala (CC-2) and R. glutinis (LMS) were ef-
fundamental research and practical applications. Antonie Van
ficient in protecting seedlings and plants; LMA1 and LMS Leeuwenhoek 99:93–105
also protected fruits; and LMA1 and CC-2 were also efficient Halfeld-Vieira BA, Silva WLM, Schurt DA, Ishida AKN, Souza GR,
for treating melon seeds. Thus, the application of these yeasts, Nechet KL (2015) Understanding the mechanism of biological con-
in combination with compatible methods of control, such as trol of passion fruit bacterial blight promoted by autochthonous
phylloplane bacteria. Biol Control 80:40–49
resistant cultivars, use of cupric compounds, and eventually Hopkins DL, Thompson CM (2002) Evaluation of Citrullus sp. germ
with ASM at a lower frequency, will be important for the plasm for resistance to Acidovorax avenae subsp. citrulli. Plant
integrated management of BFP. Dis 86:61–64
Trop. plant pathol.

IBGE (2013) Instituto Brasileiro de Geografia e Estatística (2013) Qin GZ, Tian SP, Liu HB, Xu Y (2003) Biocontrol efficacy of three
Produção Agrícola Municipal: Culturas temporárias e permanentes, antagonistic yeasts against Penicillium expansum in harvested apple
Rio de Janeiro. Available at: ftp://ftp.ibge.gov.br/Producao_ fruits. Acta Bot Sin 45:417–421
Agricola/Producao_Agricola_Municipal_[anual]/2013/tabelas_pdf/ Rosa MM, Tauk-Tornisielo SM, Rampazzo PE, Ceccato-Antonini SR
tabela02.pdf. Accessed on 26 Jan 2015 (2010) Evaluation of the biological control by the yeast
Latin R, Hopkins DL (1995) Bacterial fruit blotch of watermelon: the Torulaspora globosa against Colletotrichum sublineolum in sor-
hypothetical exam question becomes reality. Plant Dis 79:761–765 ghum. World J Microbiol Biotechnol 26:1491–1502
Lu L, Ye C, Guo S, Sheng K, Shao L, Zhou T, Yu T, Zheng X (2013) Sales Júnior R, Pontes Filho FST, Nunes GHS, Torres GRC (2007)
Preharvest application of antagonistic yeast Rhodosporidium Eficiência de acibenzolar-S-methyl e oxicloreto de cobre no controle
paludigenum induced resistance against postharvest diseases in de Acidovorax avenae subsp. citrulli, agente causal da “mancha-
mandarin orange. Biol Control 67:130–136 aquosa” no meloeiro. Revista de Biologia e Ciências da Terra 7:
McKinney RH (1923) Influence of soil temperature and moisture on 66–71
infection of wheat seedlings by Helminthosporium sativum. J Shalaby ME, El-Nady MF (2008) Application of Saccharomyces
Agric Res 6:195–218 cerevisiae as a biocontrol agent against Fusarium infection of sugar
Medeiros FHV, Moraes ISF, Silva Neto EB, Silveira EB, Mariano RLR beet plants. Acta Biologica Szegediensis 52:271–275
(2009) Management of melon bacterial blotch by plant beneficial Shaner G, Finney RE (1977) The effect of nitrogen fertilization on the
bacteria. Phytoparasitica 37:453–460 expression of slow-mildewing resistance in Knox wheat.
Mello MRF, Silveira EB, Viana IO, Guerra ML, Mariano RLR (2011) Phytopathology 67:1051–1056
Uso de antibióticos e leveduras para controle da podridão-mole em
Silveira EB, Michereff SJ, Mariano RLR (2003) Severidade da mancha-
couve-chinesa. Hortic Bras 29:78–83
aquosa em meloeiro sob diferentes condições de molhamento foliar
Oliveira IS, Sales Júnior R, Mariano RLR (2003) Ocorrência da mancha-
e concentração de inóculo de Acidovorax avenae subsp. citrulli.
aquosa causada por Acidovorax avenae subsp. citrulli, em melão-
Fitopatol Bras 28:171–175
pepino no Brasil. Fitopatol Bras 28:686
Oliveira A, Santos MHM, Silveira EB, Gomes AMA, Mariano RLR Somodi GC, Jones JB, Hopkins DL, Stall RE, Kucharek TA, Hodge NC,
(2006) Biocontrole da mancha-aquosa do melão pelo tratamento Watterson JC (1991) Occurrence of a bacterial watermelon fruit
de sementes com bactérias epifíticas e endofíticas. Hortic Bras 24: blotch in Florida. Plant Dis 75:1053–1056
373–377 Vassilev N, Vassileva M, Nikolaeva I (2006) Simultaneous P-solubilizing
Olstorpe M, Schnürer J, Passoth (2012) Growth inhibition of various and biocontrol activity of microorganisms: potentials and future
Enterobacteriaceae species by the yeast Hansenula anomala during trends. Appl Microbiol Biotechnol 71:137–144
storage of moist cereal grain. Appl Environ Microbiol 78:292–294 Walker GM (2011) Pichia anomala: cell physiology and biotechnology
Platania C, Restuccia C, Muccilli S, Cirvilleri G (2012) Efficacy of killer relative to other yeasts. Antonie Van Leeuwenhoek 99:25–34
yeasts in the biological control of Penicillium digitatum on Tarocco Wang X, Li G, Jiang DH, Huang HC (2009) Screening of plant epiphytic
orange fruits (Citrus sinensis). Food Microbiol 30:219–225 yeasts for biocontrol of bacterial fruit blotch (Acidovorax avenae
Polonelli L, Morace G (1986) Reevaluation of the yeast killer phenome- subsp. citrulli) of hami melon. Biol Control 50:164–171
non. J Clin Microbiol 24:866–869 Zhang HY, Wang SZ, Huang XY, Dong Y, Zheng XD (2008) Integrated
Pusey PL, Stockwell VO, Mazzola M (2009) Epiphytic bacteria and control of postharvest blue mold decay of pears with hot water
yeasts on apple blossoms and their potential as antagonists of treatment and Rhodotorula glutinis. Postharvest Biol Technol 49:
Erwinia amylovora. Phytopathology 99:571–581 308–313