scholarly journals Detection of Latent Monilinia Infections in Nectarine Flowers and Fruit by qPCR

Plant Disease ◽  
2017 ◽  
Vol 101 (6) ◽  
pp. 1002-1008 ◽  
Author(s):  
C. Garcia-Benitez ◽  
P. Melgarejo ◽  
A. De Cal

Most stone fruit with a latent brown rot infection caused by Monilinia do not develop visible signs of disease until the arrival of fruit at the markets or the consumer’s homes. The overnight freezing-incubation technique (ONFIT) is a well-established method for detecting latent brown rot infections, but it takes between 7 to 9 days. In this report, we inform on the advantages of applying a qPCR-based method to (i) detect a latent brown rot infection in the blossoms and fruit of nectarine trees (Prunus persica var. nucipersica) and (ii) distinguish between the Monilinia spp. in them. For applying this qPCR-based method, artificial latent infections were established in nectarine flowers and fruit using 10 Monilinia fructicola isolates, 8 M. fructigena isolates, and 10 M. laxa isolates. We detected greater amounts of M. fructicola DNA than M. laxa and M. fructigena DNA in latently infected flowers using qPCR. However, greater DNA amounts of M. laxa than M. fructicola were detected in the mesocarp of latently infected nectarines. We found that the qPCR-based method is more sensitive, reliable, and quicker than ONFIT for detecting a latent brown rot infection, and could be very useful in those countries where Monilinia spp. are classified as quarantine pathogens.

Plant Disease ◽  
2010 ◽  
Vol 94 (9) ◽  
pp. 1166-1166 ◽  
Author(s):  
A. Munda ◽  
M. Viršček Marn

Monilinia fructicola, the causal agent of brown rot, is a destructive fungal pathogen that affects mainly stone fruits (Prunoideae). It causes fruit rot, blossom wilt, twig blight, and canker formation and is common in North and South America, Australia, and New Zealand. M. fructicola is listed as a quarantine pathogen in the European Union and was absent from this region until 2001 when it was detected in France. In August 2009, mature peaches (Prunus persica cv. Royal Glory) with brown rot were found in a 5-year-old orchard in Goriška, western Slovenia. Symptoms included fruit lesions and mummified fruits. Lesions were brown, round, rapidly extending, and covered with abundant gray-to-buff conidial tufts. The pathogen was isolated in pure culture and identified based on morphological and molecular characters. Colonies on potato dextrose agar (PDA) incubated at 25°C in darkness had an average daily growth rate of 7.7 mm. They were initially colorless and later they were light gray with black stromatal plates and dense, hazel sporogenous mycelium. Colony margins were even. Sporulation was abundant and usually developed in distinct concentric zones. Limoniform conidia, produced in branched chains, measured 10.1 to 17.7 μm (mean = 12.1 μm) × 6.2 to 8.6 μm (mean = 7.3 μm) on PDA. Germinating conidia produced single germ tubes whose mean length ranged from 251 to 415 μm. Microconidia were abundant, globose, and 3 μm in diameter. Morphological characters resembled those described for M. fructicola (1). Morphological identification was confirmed by amplifying genomic DNA of isolates with M. fructicola species-specific primers (2–4). Sequence of the internal transcribed spacer (ITS) region (spanning ITS1 and ITS 2 plus 5.8 rDNA) of a representative isolate was generated using primers ITS1 and ITS4 and deposited in GenBank (Accession No. GU967379). BLAST analysis of the 516-bp PCR product revealed 100% identity with several sequences deposited for M. fructicola in NCBI GenBank. Pathogenicity was tested by inoculating five mature surface-sterilized peaches with 10 μl of a conidial suspension (104 conidia ml–1) obtained from one representative isolate. Sterile distilled water was used as a control. Peaches were wounded prior to inoculation. After 5 days of incubation at room temperature and 100% relative humidity, typical brown rot symptoms developed around the inoculation point, while controls showed no symptoms. M. fructicola was reisolated from lesion margins. Peach and nectarine orchards in a 5-km radius from the outbreak site were surveyed in September 2009 and M. fructicola was confirmed on mummified fruits from seven orchards. The pathogen was not detected in orchards from other regions of the country, where only the two endemic species M. laxa and M. fructigena were present. To our knowledge, this is the first report of M. fructicola associated with brown rot of stone fruits in Slovenia. References: (1) L. R. Batra. Page 106 in: World Species of Monilinia (Fungi): Their Ecology, Biosystematics and Control. J. Cramer, Berlin, 1991. (2) M.-J. Côté et al. Plant Dis. 88:1219, 2004. (3) K. J. D. Hughes et al. EPPO Bull. 30:507, 2000. (4) R. Ioos and P. Frey. Eur. J. Plant Pathol. 106:373, 2000.


2003 ◽  
Vol 69 (12) ◽  
pp. 7145-7152 ◽  
Author(s):  
Zhonghua Ma ◽  
Michael A. Yoshimura ◽  
Themis J. Michailides

ABSTRACT Low and high levels of resistance to the benzimidazole fungicides benomyl and thiophanate-methyl were observed in field isolates of Monilinia fructicola, which is the causative agent of brown rot of stone fruit. Isolates that had low levels of resistance (hereafter referred to as LR isolates) and high levels of resistance (hereafter referred to as HR isolates) were also cold and heat sensitive, respectively. Results from microsatellite DNA fingerprints showed that genetic identities among the populations of sensitive (S), LR, and HR isolates were very high (>0.96). Analysis of DNA sequences of theβ -tubulin gene showed that the LR isolates had a point mutation at codon 6, causing a replacement of the amino acid histidine by tyrosine. Codon 198, which encodes a glutamic acid in S and LR isolates, was converted to a codon for alanine in HR isolates. Based on these point mutations in the β-tubulin gene, allele-specific PCR assays were developed for rapid detection of benzimidazole-resistant isolates of M. fructicola from stone fruit.


Plant Disease ◽  
2010 ◽  
Vol 94 (1) ◽  
pp. 126-126 ◽  
Author(s):  
A. Burnett ◽  
N. Lalancette ◽  
K. McFarland

Reduced sensitivity and resistance of Monilinia fructicola to demethylation inhibitors (DMIs; fungicide group 3) have been previously found in stone fruit orchards in Georgia, South Carolina, Ohio, and New York (2). Resistance development is a major concern because of the importance of DMIs for brown rot management. Eleven single-spore isolates, originally collected during 2006 from separate commercial peach (Prunus persica) orchards in southern New Jersey, were removed from cold storage (5°C) in early 2008 and examined in vitro for resistance to the DMI propiconazole (Orbit 3.6EC; Syngenta Crop Protection, Inc., Greensboro, NC). After 19 months at 5°C, isolate 7 was inhibited 53.4% in growth on potato dextrose agar (PDA) amended at the discretionary dose of 0.3 μg/ml propiconazole; inhibition of the remaining isolates ranged from 81.4 to 100%. Inhibition values were based on two replications of eight colonies per isolate performed after incubation at 25°C for 4 days. Because of the previously reported relationship between duration of cold storage and propiconazole sensitivity, isolate 7 was tentatively deemed resistant (1). To confirm the in vitro results, isolates were grown at 25°C for 7 days on cellophane over PDA. Genomic DNA was isolated from mycelium with the DNeasy Plant Mini Kit (Qiagen, Inc., Valencia, CA). PCR with primers INS65-F and INS65-R was conducted on a GeneAmp thermal cycler (Applied Biosystems, Inc., Foster City, CA) as described previously to amplify a 65-bp region named ‘Mona’ associated with DMI resistance (2). PCR products were separated via electrophoresis on 0.8% agarose gel. The primers amplified a 376-bp fragment from isolate 7 and a 311-bp fragment from all other isolates, thus indicating the presence of Mona in isolate 7. Restriction fragment length polymorphism analysis using the BsrBI enzyme, specific to a single restriction site within Mona, was conducted on the amplified fragments from all isolates. Electrophoresis results showed digestion of the 376-bp fragment from isolate 7 into 140-bp and 236-bp fragments, thereby confirming the presence of Mona; none of the 311-bp fragments from the remaining isolates were cut by BsrBI. Although economic loss from brown rot has not been reported in New Jersey, these results show that propiconazole-resistant strains have been detected since 2006 and it is most likely that resistant strains of the pathogen are still present in commercial peach orchards. To combat this risk, current brown rot control recommendations are incorporating quinone outside inhibitors (QoIs; fungicide group 11) and carboxamides (fungicide group 7) into control programs as a resistance management strategy. More extensive sampling is planned to ascertain the prevalence and location of resistant strains. References: (1) K. D. Cox et al. Phytopathology 97:448, 2007. (2) C.-X. Luo et al. Plant Dis. 92:1099, 2008.


Plant Disease ◽  
2020 ◽  
Vol 104 (11) ◽  
pp. 2843-2850
Author(s):  
Pamela Suellen Salvador Dutra ◽  
Paulo S. F. Lichtemberg ◽  
Maria Bernat Martinez ◽  
Themis J. Michailides ◽  
Louise Larissa May De Mio

Despite the resistance problems in Monilinia fructicola, demethylation inhibitor fungicides (DMIs) are still effective for the disease management of brown rot in commercial stone fruit orchards in Brazil. This study aims to investigate the sensitivity of M. fructicola isolates and efficiency of DMIs to reduce brown rot. A set of 93 isolates collected from Brazilian commercial orchards were tested for their sensitivities to tebuconazole, propiconazole, prothioconazole, and myclobutanil. The isolates were analyzed separately according to the presence or absence of the G461S mutation in MfCYP51 gene, determined by allele-specific test. The mean EC50 values for G461S mutants and wild-type isolates were respectively 8.443 and 1.13 µg/ml for myclobutanil, 0.236 and 0.026 µg/ml for propiconazole, 0.115 and 0.002 µg/ml for prothioconazole, and 1.482 and 0.096 µg/ml for tebuconazole. The density distribution curves of DMI sensitivity for both genotypes showed that myclobutanil and prothioconazole curves were mostly shifted toward resistance and sensitivity, respectively. Incomplete cross-resistance was detected among propiconazole and tebuconazole in both wild-type (r = 0.45) and G461S (r = 0.38) populations. No cross-sensitivity was observed among wild-type isolates to prothioconazole and the others DMIs tested. Fungicide treatments on detached fruit inoculated with M. fructicola genotypes showed significant DMI efficacy differences when fruit were inoculated with wild-type and G461S isolates. Protective applications with prothioconazole were more effective for control of both G461S and wild-type isolates compared with tebuconazole. Curative applications with tebuconazole were most effective in reducing the incidence and lesion size of G461S isolates. Sporulation occurred only for G461S isolates treated with tebuconazole under curative and preventative treatments. The differences found among the performance of triazoles against M. fructicola isolates will form the basis for recommendations of rational DMI usage to control brown rot in Brazil.


Plant Disease ◽  
2013 ◽  
Vol 97 (9) ◽  
pp. 1255-1255 ◽  
Author(s):  
L. F. Yin ◽  
S. N. Chen ◽  
N. N. Yuan ◽  
L. X. Zhai ◽  
G. Q. Li ◽  
...  

Brown rot of peach (Prunus persica) in China has been reported to be caused by at least three Monilinia species (1). In the present study, peaches with symptoms resembling brown rot caused by Monilinia species were collected from commercial orchards in the northwestern province of Gansu in August 2010, the southwestern province of Yunnan in July 2011, and in the central province of Hubei in July 2012. Affected fruit showed the typical symptoms of brown rot with zones of sporulation. Fungal isolates were single-spored and cultured on potato dextrose agar (PDA). Colonies showed grayness with concentric rings of sporulation after incubation at 25°C in the dark. Mean mycelial growth of isolates YHC11-1a and YHC11-2a from Yunnan, GTC10-1a and GTC10-2a from Gansu, and HWC12-14a and HWC12-23a from Hubei, was 4.6 ± 0.4 and 7.5 ± 0.7 cm after 3 and 5 days incubation, respectively. Conidia were lemon shaped and formed in branched monilioid chains, and the mean size was 9.3 (6.7 to 11.5) × 12.5 (7.9 to 17.8) μm, which was consistent with the characteristics of Monilinia fructicola (1,2). The species identification was confirmed by sequencing of the ribosomal ITS sequences. The ribosomal ITS1-5.8S-ITS2 region was amplified from each of the six isolates using primers ITS1 and ITS4 (3). Results indicated that the ITS sequences of these isolates were identical and showed the highest similarity (100%) with M. fructicola ITS sequences from isolates collected in China (GenBank Accession Nos. HQ893748, FJ515894, and AM887528), Slovenia (GU967379), Italy (FJ411109), and Spain (EF207423). The pathogen was also confirmed to be M. fructicola based on the detection of an M. fructicola- specific band (534 bp) using a PCR-based molecular tool developed for distinguishing Chinese Monilinia species affecting peach (1). Pathogenicity was tested on surface-sterilized, mature peaches (Shui Mi Tao) with representative isolates. Fruits were holed at three equidistant positions to a depth of 5 mm using a sterile cork borer. Mycelial plugs (5 mm in diameter) from the periphery of a 4-day-old colony of each isolate were placed upside down into each hole, control fruits received water agar. After 3 days of incubation at 22°C in a moist chamber, inoculated fruits developed typical brown rot symptoms while control fruits remained healthy. Pathogens from the inoculated fruit were confirmed to be M. fructicola based on morphological characteristics. To our knowledge, this is the first report of occurrence of M. fructicola in Gansu, Yunnan, and Hubei provinces, thousands of kilometers away from eastern China where occurrence of peach brown rot caused by M. fructicola has been confirmed (2,4). The results indicated the further geographical spread of the M. fructicola in China. References: (1) M. J. Hu et al. Plos One 6(9):e24990, 2011. (2) M. J. Hu et al. Plant Dis. 95:225, 2011. (3) T. J. White et al. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. Academic Press, San Diego, 1990. (4) X. Q. Zhu et al. Plant Pathol. 54:575, 2005.


Plant Disease ◽  
2001 ◽  
Vol 85 (9) ◽  
pp. 999-1003 ◽  
Author(s):  
Yong Luo ◽  
Zhonghua Ma ◽  
Themis J. Michailides

Two studies were conducted to determine the effects of water content (WC) on sporulation on thinned fruit and the effects of wetness duration, inoculum density, and temperature on secondary infection of prune fruit by Monilinia fructicola, the main causal pathogen of brown rot in California. In the first study, sporulation intensity and duration of sporulation of the pathogen were tested on inoculated thinned fruit with five levels (67.2, 53.8, 40.3, 26.9, and 13.4%) of WC. Regression analyses showed that both sporulation intensity and duration of sporulation increased as WC of thinned fruit increased. The predicted difference in duration of sporulation between fruit with 13.4 and 67.2% WC was about 3 days. In the second study, three inoculum concentrations (8,000, 16,000, and 24,000 conidia per milliliter) of M. fructicola were atomized onto prune fruit on trees in an orchard. Inoculated fruit and shoots were covered with plastic bags to maintain wetness duration for 4, 8, 12, or 16 h. An overnight freezing and incubation technique was used after harvest to determine the proportion of fruit with latent infection. Regression analysis demonstrated that inoculum concentration and wetness duration were significant factors affecting secondary infection. Temperature was less important. Increased inoculum concentration and wetness duration increased the percentage of fruit with latent infections. Increased temperature decreased the percentage of fruit with latent infections.


Plant Disease ◽  
2020 ◽  
Author(s):  
Kim Lesniak ◽  
Jingyu Peng ◽  
Tyre J Proffer ◽  
Cory Outwater ◽  
Lauren Eldred ◽  
...  

Resistance to sterol demethylation inhibitor fungicides (DMIs) in Monilinia fructicola, causal agent of brown rot of stone fruit, has been reported in the southeastern and eastern United States and in Brazil. DMI resistance of some M. fructicola isolates, in particular those recovered from the southeastern U.S., is associated with a sequence element termed ‘Mona’ that causes overexpression of the cytochrome demethylase target gene MfCYP51. In this study, we conducted statewide surveys of Michigan stone fruit orchards from 2009-2011 and in 2019, and determined the sensitivity to propiconazole of a total of 813 isolates of M. fructicola. A total of 80.7% of Michigan isolates were characterized as resistant to propiconazole by relative growth assays but the ‘Mona’ insert was not uniformly detected, and was present in some isolates that were not characterized as DMI resistant. Gene expression assays indicated that elevated expression of MfCYP51 was only weakly correlated with DMI-resistance in M. fructicola isolates from Michigan, and there was no obvious correlation between the presence of the ‘Mona’ element and elevated expression of MfCYP51. However, sequence analysis of MfCYP51 from 25 DMI-resistant isolates did not reveal any point mutations that could be correlated with resistance. Amplification and sequencing upstream of MfCYP51 resulted in detection of DNA insertions in a wide range of isolates typed by DMI phenotype and the presence of ‘Mona’ or other unique sequences. The function of these unique sequences or their presence upstream of MfCYP51 cannot be correlated to a DMI-resistant genotype at this time. Our results indicate that DMI resistance was established in Michigan populations of M. fructicola by 2009 to 2011, and that relative resistance levels have continued to increase to the point that practical resistance is present in most orchards. In addition, the presence of the ‘Mona’ insert is not a marker for identifying DMI-resistant isolates of M. fructicola in Michigan.


Plant Disease ◽  
1997 ◽  
Vol 81 (5) ◽  
pp. 519-524 ◽  
Author(s):  
Chuanxue Hong ◽  
Brent A. Holtz ◽  
David P. Morgan ◽  
Themis J. Michailides

The significance of thinned fruit as a source of secondary inoculum in the spread of brown rot, caused by Monilinia fructicola, under semi-arid weather conditions of the San Joaquin Valley in California, was investigated in seven nectarine orchards in 1995 and 1996. Between 6 and 60% (depending on the orchard) of thinned fruit showed sporulation by M. fructicola. Brown rot was significantly less severe at preharvest (five orchards) and postharvest (one orchard) on fruit harvested from trees in plots from which thinned fruit were completely removed than on those in plots from which thinned fruit were not removed. M. fructicola sporulated more frequently on thinned fruit placed into irrigation trenches than on those left on the dry berms in tree rows. The incidence of preharvest fruit brown rot increased exponentially as the density of thinned fruit increased on the orchard floor. These results suggest that thinned fruit left on the floor of nectarine orchards can be a significant inoculum source of secondary infections. Removal or destruction of thinned fruit should reduce brown rot in nectarine and possibly other stone fruit orchards under semi-arid California conditions.


Plant Disease ◽  
2007 ◽  
Vol 91 (7) ◽  
pp. 907-907 ◽  
Author(s):  
J. Duchoslavová ◽  
I. Širučková ◽  
E. Zapletalová ◽  
M. Navrátil ◽  
D. Šafářová

Monilinia fructicola, a causal agent of brown rot, is one of the most important fungal pathogens of stone fruits. The disease causes major crop losses in peach, plum, prune, nectarine, and apricot. M. fructicola is commonly present in Asia, North and South America, and Australia. This is a quarantined pathogen in Europe; restricted occurrence has been observed in Austria and France. Recently, it was detected in Hungary and Switzerland on peach and nectarine fruits imported from Italy and Spain (1,4). During a survey in the summer of 2006, 56 samples were tested for the presence of Monilinia spp. M. fructicola was detected in 15 samples from 11 locations in the western area (Bohemia) of the Czech Republic, mainly on peaches (Prunus persica), apples (Malus × domestica), and sweet and sour cherries (Cerasus avium and C. vulgaris) and rarely on flowering plum (Prunus triloba) and Malus × moerlandsii cv. Liset. On the other hand, the pathogen was not detected on fruits of apricot (Prunus armeniaca) or pear (Pyrus communis). In all cases, M. fructicola was detected on fruits except for a single occurrence of the pathogen on a shoot of the Malus × domestica. The pathogen was always detected in mixed infections with M. fructigena and/or M. laxa. On both fruits and the shoot, symptoms appeared as brown, sunken lesions covered with grayish pustules. Many infected fruits became dry and mummified because rot progressed through the fruit surface. The infected shoot died back (3). M. fructicola was identified by means of colony and conidial morphology and molecular characteristics. The colonies cultivated on potato dextrose agar were entire and the colony surface was even. The color of the colony was gray, and sporulating colonies showed concentric rings that changed to a hazel color. Conidia were ellipsoid, hyaline, and 13.5 to 17.7 × 8.3 to 10.5 μm. Preliminary morphological identification was confirmed by PCR (2) on DNA isolated directly from mycelium on the examined fruits. A product that was 280 bp long was obtained in all cases. The BLAST analysis of our PCR product sequences showed 100% homology to sequences of M. fructicola (GenBank Accession Nos. DQ491506, AY2891185, Z73778, and AB125615). One sequence from our study was deposited in GenBank (Accession No. EF378628). To our knowledge, this is the first report of the quarantined fungus M. fructicola in the Czech Republic. References: (1) E. Bosshard et al. Plant Dis. 90:1554, 2006. (2) K. J. D. Hughes et al. EPPO Bull. 30:507, 2000. (3) J. M. Ogawa et al., eds. Compendium of Stone Fruit Diseases. The American Phytopathological Society, St. Paul, MN, 1995. (4) M. Petróczy and L. Palkovics. Plant Dis. 90:375, 2006.


Plant Disease ◽  
2009 ◽  
Vol 93 (6) ◽  
pp. 668-668 ◽  
Author(s):  
C. Pellegrino ◽  
M. L. Gullino ◽  
A. Garibaldi ◽  
D. Spadaro

Monilinia fructicola, causal agent of brown rot, is one of the most important fungal pathogens of stone fruit. M. fructicola is a quarantined pathogen in Europe. During the summer of 2008 in 15 orchards located in Piedmont (northern Italy), 12,500 stone fruits (cherries, apricots, peaches, nectarines, and plums) were stored in cold chambers at 4 and 6°C and monitored for 8 weeks for the presence of Monilinia spp. M. fructicola was detected on 0.5% of nectarines (cvs. Sweet Red and Orion) that originated from two orchards in Lagnasco. Symptoms appeared on the fruit during storage, starting 3 weeks after harvest. Fruit rot lesions were brown, sunken, and covered with grayish tufts. The majority of infected fruit became dry and mummified. Brown rot symptoms were similar to those caused by endemic M. fructigena and M. laxa. Symptoms began with a small, circular, brown spot, and the rot spread rapidly. At the same time, numerous, small, grayish stromata developed. Finally, the whole surface of the fruit was covered by conidial tufts. Tissues were excised from diseased stone fruits and cultured on potato dextrose agar (PDA) amended with 25 μg of streptomycin per liter. The isolates produced abundant mycelium on PDA at 20 ± 2°C. Colonies were initially gray, but after sporulation, they became hazel, showing concentric rings (sporulation is sparse in M. laxa or M. fructigena). Conidia were one-celled, ellipsoid, hyaline, 15.2 × 10.1 μm, and produced in branched monilioid chains (2). Preliminary morphological identification of fungi resembling M. fructicola was confirmed by PCR using genomic DNA extracted from the mycelia of pure cultures. The DNA was amplified with a common reverse primer and three species-specific forward primers (3) obtained from a sequence characterized amplified region and a product of 535 bp, diagnostic for the species M. fructicola, was obtained. BLAST analysis of the amplified sequence (GenBank Accession No. FI569728) showed 96% similarity to the sequence of a M. fructicola isolated from Canada (GenBank Accession No. AF506700), 15% similarity to M. laxa ATCC11790 (GenBank Accession No. AF506702), and 35% similarity to a M. fructigena sequence isolated in Italy (GenBank Accession No. AF506701). Moreover, two sequences obtained through the amplification of ribosomal region ITS1-5.8S-ITS2, showing 100% similarity to the same ribosomal sequence of M. fructicola, were deposited in GenBank (Accession Nos. FJ411109 and FJ411110). The pathogen was detected on some mummified fruit from the same orchards in November of 2008. Pathogenicity was tested by spraying 103 conidia/ml on 10 surface-sterilized artificially wounded nectarines per strain of M. fructicola. After 5 days of incubation at 20 ± 2°C, typical, brown, rot symptoms developed on inoculated fruit. M. fructicola was reisolated from the inoculated fruit on PDA. Symptoms did not appear on control fruit. To our knowledge, this is the first report of M. fructicola in Italy. Its occurrence in Europe has been reported sporadically in Austria and France, and in 2006, it was detected in Hungary and Switzerland on peaches and nectarines imported from Italy and Spain (1,4). References: (1) E. Bosshard et al. Plant Dis. 90:1554, 2006. (2) R. J. W. Byrde and H. J. Willetts. The Brown Rot Fungi of Fruit: Their Biology and Control. Pergamon Press, Oxford, 1977. (3) M. J. Coté et al. Plant Dis. 88:1219, 2004. (4) M. Petròczy and L. Palkovics. Plant Dis. 90:375, 2006.


Sign in / Sign up

Export Citation Format

Share Document